A man with metastatic prostate cancer has exhausted standard options. His oncologist wants to give lutetium-177-PSMA-617. The temptation is to treat on diagnosis alone — it is prostate cancer, the drug is for prostate cancer, give it. But first, they image. Several large, symptomatic lesions show almost no PSMA signal. Those lesions have lost the target. Had the team treated on diagnosis alone, they would have irradiated salivary glands and kidneys while the tumors that mattered received almost nothing. The scan is not a formality. It is the patient-selection step.

Theranostics combines therapy and diagnostics in one molecular framework. The imaging agent and the therapeutic agent target the same molecule. In conventional oncology, imaging and treatment are separate — a patient is staged with one agent and treated with an unrelated drug. In theranostics, the loop closes. The image becomes a genuine companion diagnostic for that specific therapy, not a generic staging procedure. The question the scan asks is exactly the question the therapy requires an answer to: does this patient's tumor express the target?

Prostate-specific membrane antigen is expressed at high levels on prostate cancer cells, and also on salivary glands and renal tubules — the dose-limiting normal tissues. The imaging side uses small molecules labeled with PET isotopes: gallium-68-PSMA-11 and fluorine-18-DCFPyL are both FDA-approved. The therapy side loads the same targeting molecule with lutetium-177. Lutetium-177-PSMA-617, brand name Pluvicto, was approved in 2022 for PSMA-positive metastatic castration-resistant prostate cancer. The VISION trial showed improved overall survival in heavily pretreated patients. Because imaging and therapy share the targeting molecule, the PSMA PET scan is not regulatory caution — it is mechanistically mandatory.

The second established pair follows the same logic in a different cancer. Lutetium-177-DOTATATE targets somatostatin receptor type 2, overexpressed on gastroenteropancreatic neuroendocrine tumors. Its companion imaging agent, gallium-68-DOTATATE PET, images the same receptor. DOTATATE was the first FDA-approved radioligand theranostic in the modern era, approved in 2018, and established the regulatory pathway that PSMA-617 followed four years later. Same target, two isotopes, imaging before therapy — the architecture is identical. One loop, applied to a completely different molecular handle.

Beta emitters release electrons. Lutetium-177 is the most clinically used; yttrium-90 has higher energy and longer range; iodine-131 is the classic thyroid isotope. The key physical property is range: beta particles travel several millimeters through tissue. This produces crossfire — a cell that does not express the target can still be killed by radiation from a neighboring cell that has bound the radioligand. Crossfire is directly useful in tumors with heterogeneous target expression. The cost is lower linear energy transfer — beta particles deposit energy sparsely, and cells can sometimes repair the damage.

PSMA is expressed not only on prostate cancer cells but on salivary glands and renal tubules. The radioligand binds wherever the target is present. Regardless of how clearly the tumor lights up on PET, the delivered dose is shared between tumor and normal organs, and the tolerances of those normal organs cap how much isotope can be given. The predictable toxicities follow directly: xerostomia from salivary gland uptake, kidney toxicity from renal tubule uptake, and bone marrow suppression. You cannot simply give more isotope to deliver a higher tumor dose. More isotope also increases the salivary and renal dose, and those organs have tolerance limits crossed before the tumor dose reaches what might be needed to overcome resistance.

Absorbed dose — energy deposited per unit mass of tissue, measured in gray — is the quantity that connects radioactivity to biological effect. The therapeutic goal is to deliver a lethal absorbed dose to target-expressing cells. The constraint is keeping the absorbed dose to salivary glands, kidneys, and bone marrow below their tolerances. The gap between the two is the therapeutic window, and it is narrower than the simplest version of the theranostic story suggests. This is also exactly why the opening patient with PSMA-negative metastases would have been harmed rather than helped: his salivary glands and kidneys would have received their full dose, while the target-negative tumors were largely spared.

Radioligand theranostics has several approved therapies with clinical benefit demonstrated in randomized trials. Engineered nanoparticles with multiple functions — targeting ligands, drug payloads, imaging labels, release triggers — have appeared promising in cell culture and animal models but have largely failed to translate. Why the difference? The answer is in simplicity. A radioligand is a small molecule carrying an isotope. No nanoparticle-scale architecture to characterize, no protein corona, no dose-loss chain between injection and intracellular payload release. It circulates, binds its target, and emits. The design succeeds because it is mechanistically parsimonious: one target, one binding event, two isotopes.

Here is the chapter's central claim. Radioligand theranostics succeeds because its companion-diagnostic loop is built from a shared targeting molecule — making patient selection mechanistic rather than statistical. And emitter choice must be matched to tumor geometry: alpha emitters for micrometastatic or uniformly positive disease, beta emitters for heterogeneous bulky tumors where crossfire is needed. Two findings would force revision. If PSMA-negative patients showed comparable survival benefit, the imaging step would be confirmed as staging rather than companion diagnostic. If alpha emitters outperformed beta even in demonstrably heterogeneous tumors, the geometry-matching argument would need revision. Both questions are under active investigation.