Molecular imaging

Molecular imaging is the diagnostic half of theranostics: a radiolabelled ligand binds a target expressed on tumour cells (PSMA, SSTR2, FAP, HER2, CAIX, TROP2 and others) and a PET/CT or PET/MRI camera maps its distribution across the whole body. The field's biggest single deployment is PSMA PET in prostate cancer: the joint EANM/SNMMI PSMA PET/CT procedure guideline 2.0 (Fendler et al., Eur J Nucl Med Mol Imaging 2023) and the 2026 NCCN guidelines now recommend PSMA PET ahead of conventional imaging for unfavourable intermediate- and high-risk staging and for biochemical recurrence, and require it to select patients for PSMA-targeted radioligand therapy. The 2020 proPSMA trial (Hofman et al., Lancet) established the head-to-head case against CT and bone scan.

Somatostatin receptor (SSTR) PET with 68Ga-DOTATATE or DOTATOC plays the equivalent gatekeeping role for peptide receptor radionuclide therapy in neuroendocrine tumours, and FAPI PET is being explored as a pan-cancer stromal imaging agent, though it does not yet have the guideline weight of PSMA or SSTR PET. FDG PET remains the workhorse general-oncology tracer and the comparator against which every newer agent is judged.

What is unsolved in 2026: PSMA expression is heterogeneous and a meaningful fraction of mCRPC patients are PSMA-negative or discordant on biopsy, so imaging-based patient selection is an imperfect proxy for radioligand therapy response; quantitative reporting (SUV, PROMISE/miTNM, E-PSMA) is still not uniformly applied across sites; and reimbursement for newer PET tracers lags their guideline recommendation in several health systems. Total-body PET and AI-assisted reading are being investigated to cut dose and scan time but are not yet standard of care.