FAP-targeted radioligand therapy for sarcoma

Fibroblast activation protein (FAP) is usually found on the stromal cells surrounding a tumour rather than on cancer cells directly, which limits how selective a FAP-targeted therapy can be, but sarcomas are one of the few cancer types where the tumour cells themselves can express FAP. 68Ga-FAPI PET has shown superior lesion detection compared with 18F-FDG PET across multiple sarcoma subtypes in prospective imaging cohorts, making FAP an attractive target for a therapeutic radioligand in a disease with essentially no PSMA or somatostatin-receptor expression to exploit.

The handful of early-phase therapeutic studies published so far give a mixed signal. A 5-patient pilot study of 177Lu-FAPI-2286 in metastatic sarcoma reported partial response in 4 of 5 patients by RECIST, with a 52% mean reduction in primary tumour volume. A separate, later 6-patient exploratory study of the same drug in metastatic sarcoma found no RECIST partial or complete responses at all, only stable disease in some patients (with mixed results by other volumetric metrics), a median overall survival of just 4 months, and 3 of 6 patients dying during follow-up. A first-in-human dose-escalation trial of a related FAP ligand, 177Lu-FAPI-XT (NCT06211647/NCT06197139, 14 patients across several tumour types including fibroblastic sarcoma subtypes), found an overall disease control rate of 35.7%, rising to 50% in the sarcoma subtypes specifically, but no complete or partial responses in the whole cohort, alongside a favourable safety profile with no dose-limiting toxicity up to 11.1 GBq. Separately, the Novartis-sponsored LuMIERE trial (NCT04939610) is testing 177Lu-FAP-2286 in pancreatic, non-small-cell lung and breast cancer, not specifically in sarcoma.

These are all small, uncontrolled, single-arm case series (5-14 patients each) with conflicting response rates, so whether FAP-targeted radioligand therapy has genuine, reproducible activity in sarcoma, and if so in which histological subtypes, remains unresolved.