Radioligand Target Due Diligence: A 12-Point Checklist for BD and Investment Teams

Updated October 2026.

Radiopharmaceutical deal flow has outpaced the number of people who can evaluate it. Many in-licensing and investment decisions are made by teams with deep oncology or antibody experience but limited radiopharmacy background. Radioligands fail for reasons that look different from ADC or small-molecule failures: dose to the kidney and salivary glands, daughter isotopes, shedding, vector pharmacokinetics.

This checklist is the set of questions we would want answered about the target before looking at the asset's data package. Each point includes the signal that should raise concern.

Biology: can a radioligand reach it, and does it stay?

1. Is the target actually on the cell surface? Check localization from more than one source, such as UniProt, the Human Protein Atlas subcellular atlas and the primary literature. Many proteins annotated as "membrane" sit in the ER, Golgi, mitochondria or an intracellular vesicle, where a radioligand can't reach them. Red flag: surface localization supported by only one database annotation.

2. How much of the protein is outside the cell? Multi-pass proteins with small extracellular loops limit binder options and often force an antibody vector, which then constrains isotope choice and marrow dose. Red flag: no validated binder to an extracellular epitope.

3. Does it internalize after binding? Internalization retains activity in the tumor. It is close to essential for ²²⁵Ac, because recoiling daughter isotopes can escape from cells that don't internalize. See alpha vs beta target selection. Red flag: internalization asserted but shown only with a different binder or epitope than the asset's.

4. Does it shed? A soluble form of the target in plasma acts as a decoy: it binds the radioligand in the circulation and lowers tumor uptake. Red flag: known soluble isoform or protease-cleavage site, with no serum measurement in the data package.

Tumor: is it there, in enough patients, in enough cells?

5. What fraction of patients is positive? Check patient-level positivity in the target indication, not just mean expression. Look for IHC across a reasonable number of tumors and, ideally, imaging data. Red flag: expression shown only in cell lines or xenografts.

6. How homogeneous is the expression within each tumor? Beta emitters tolerate patchy expression through crossfire. Alpha emitters don't. Red flag: an alpha payload against a target with known intratumoral heterogeneity.

7. Is expression maintained in the patients who will actually be treated? Late-line, pre-treated and metastatic disease can differ from primary tumors. For example, PSMA is lost in a subset of advanced prostate cancers (see PSMA-low prostate cancer targets). Red flag: expression data only from primary, treatment-naive tissue.

Safety: where else does the dose go?

8. What is the expression in dose-limiting organs? Kidney, salivary glands, bone marrow, liver and small intestine. Any expression there turns directly into absorbed dose. See normal-tissue expression and dosimetry risk. Red flag: normal-tissue data limited to a single RNA dataset, with no protein data.

9. Is there dosimetry from imaging? A theranostic pair (an imaging analog of the therapeutic) lets you measure organ doses in patients instead of extrapolating from mice. Red flag: human doses projected only from rodent biodistribution.

Strategy: is the opportunity defensible?

10. Does the isotope fit the target and the vector? Vector half-life should match isotope half-life, and the decay physics should match the expression pattern. Red flag: a long-circulating antibody paired with an isotope chosen for supply reasons rather than fit.

11. What clinical precedent exists in any modality? Approved ADCs, T-cell engagers or antibodies against the same target de-risk expression and on-target toxicity, but they also mean competition for the same patients. Red flag: the target's only clinical precedent is a failed program, with no explanation of why radioligands would succeed.

12. How crowded is the target? Count active radioligand and non-radioligand programs, and how far along they are. A validated target with ten programs is a different bet from a de-risked target with two. Red flag: differentiation that rests only on the isotope.

Running the checklist at scale

Answering these twelve questions for one target takes a few days of literature and database work. Doing it across a full competitive landscape, or every target in an indication, is where teams run out of time.

Nuclens pre-computes the first-pass evidence for points 1, 3, 4, 5 and 11 across about 15,000 oncology targets: cell-surface localization, internalization and shedding evidence extracted from the literature, tumor IHC positivity by cancer type, and clinical stage, each linked to its source. Describe the asset or the indication, and get a ranked shortlist with the evidence attached.

Run a due-diligence screen free →

Nuclens is a first-pass triage layer. It doesn't replace expert review, wet-lab validation or patient dosimetry.