Nine peptide-receptor systems are used or proposed for in vivo tumor imaging: RGD peptides and integrin αVβ3, bombesin and gastrin-releasing peptide receptors, somatostatin and its five receptor subtypes, VIP, CCK and gastrin, α-MSH and MC-1R, neurotensin, the CXCR4 antagonist T140, and exendin-4…
In vivo tumor imaging with radiolabeled peptides rests on a density difference. A peptide that binds a cell-surface receptor is tagged with a radionuclide and injected; where the receptor is overexpressed, the signal accumulates. Nine peptide-receptor systems recur in this field: RGD peptides and integrin αVβ3, bombesin and gastrin-releasing peptide GRP analogs acting through four bombesin-like receptors, somatostatin and its synthetic analogs acting through five somatostatin receptor subtypes, vasoactive intestinal peptide VIP acting through VPAC1 and VPAC2, cholecystokinin CCK and gastrin acting through the CCK2 receptor, α-melanocyte-stimulating hormone α-MSH acting through MC-1R, neurotensin and its receptors, the CXCR4 antagonist T140, and exendin-4 acting through the GLP-1 receptor.
| System | Receptor target | Tumor types cited |
|---|---|---|
| RGD peptide | Integrins, primarily αVβ3 | Angiogenic tumor vasculature; germ cell tumor residual masses NCT02317393 |
| Bombesin / GRP | Neuromedin B receptor, bombesin receptor subtype 3, GRPR, bombesin receptor subtype 4 | Breast cancer NCT04746638 ; ovarian cancer and gastrointestinal stromal tumors; prostate cancer NCT04746638, NCT05633160 |
| Somatostatin | SSTR1 to SSTR5 | Pancreatic neuroendocrine tumors, small cell lung cancer, thyroid carcinoma |
| VIP | VPAC1, VPAC2 | Brain tumors, pancreatic adenocarcinoma, neuroendocrine tumors |
| CCK / gastrin | CCK2 / gastrin receptor | Ovarian stromal tumors, astrocytomas |
| α-MSH | MC-1R | Melanoma; MC-1R in over 80% of metastases |
| Neurotensin | Neurotensin receptors | Ductal pancreatic adenocarcinoma, small cell lung carcinoma, medullary thyroid carcinoma |
| T140 | CXCR4 | Not specified in source |
| Exendin-4 | GLP-1 receptor | Not specified source text truncated |
The imaging rationale is receptor density. The peptides listed here bind receptors that are overexpressed on tumor cells or tumor-associated tissue but present at low levels on most normal cells, and that gradient is what produces a detectable image. Several molecular properties make short peptides unusually well suited to this job. They are small, from 3 to 39 amino acids, with molecular weights from roughly 346 to 3327 daltons. They clear rapidly from the blood, which keeps background signal low. They can be assembled by solid-phase synthesis, and they tolerate the conjugation of chelators and radionuclides. The registered studies for these systems use 99mTc for single-photon emission computed tomography SPECT and 18F, 68Ga, and 64Cu for positron emission tomography PET .
One caution applies to the whole roster. The claims assembled for this review originate in a commercial peptide catalog, not a systematic review, and assurances of clinical utility attached to individual peptides should be treated as vendor framing until confirmed in the peer-reviewed or registry record. The sections below separate catalog claims from indexed literature and registered trials.
| Peptide | Amino acids | Average molecular weight | Molecular formula |
|---|---|---|---|
| RGD | 3 | 346.34 | C12H22N6O6 |
| Somatostatin | 14 or 28 | Not specified | Not specified |
| VIP | 28 | 3326.78 | C147H237N43O43S1 |
| CCK / gastrin | 8 | 1144.25 | C49H61N9O17S3 |
| α-MSH | 13 | 1623.83 | C75H106N20O19S1 |
| Neurotensin | 13 in description, 12 in data section | 1672.92 | C78H121N21O20 |
| T140 | 14 | 2039.44 | C90H143N33O18S2 |
| Exendin-4 | 39 | Not specified | Not specified |
Two features of the table deserve attention before the individual systems are discussed. The single sulfur atom in the VIP formula C147H237N43O43S1 is the methionine in the 28-residue sequence, and the same applies to the one sulfur in α-MSH. The CCK/gastrin formula C49H61N9O17S3, by contrast, carries three sulfurs, one of them the sulfate ester on the tyrosine whose position distinguishes gastrin from CCK. T140's formula C90H143N33O18S2 reflects a different chemistry: two sulfurs joined in the peptide's single disulfide bond, which constrains the molecule into the conformation it needs to block CXCR4. The amidated C-terminus of α-MSH, preserved in the common reagent form α-MSH amide, is part of the peptide's biology.
The neurotensin row shows a real discrepancy: the description says 13 amino acids, the data section says 12. That inconsistency, in a commercial listing, is a concrete reason to verify any peptide reagent's identity by mass spectrometry before it goes into an animal or a patient.
RGD is the minimal recognition sequence arginine-glycine-aspartic acid: three amino acids, an average molecular weight of 346.34, and a molecular formula of C12H22N6O6. It binds integrins, with the strongest imaging interest in the αVβ3 heterodimer, a receptor overexpressed on endothelial cells during tumor angiogenesis and involved in tumor metastasis. Because angiogenesis accompanies nearly all solid tumor growth, RGD-based tracers image the tumor vasculature rather than the tumor cell itself. Three of the five registered studies examined here target this system, more than any other in the set.
The clinical record for RGD imaging is mixed, and the pattern is instructive. NCT01176500, a pilot study of 18F-fluciclatide PET/CT for evaluating anti-angiogenic therapy in colon, lung, and cervical cancer, was withdrawn before enrolling any patients. NCT02317393, a phase 2 study, completed with 16 patients and used αVβ3 imaging to characterize residual masses of non-seminomatous germ cell tumors at the end of chemotherapy. NCT05543317, completed with 27 patients across various solid tumors, used a 68Ga-labeled dual tracer targeting both integrin αVβ3 and fibroblast activation protein FAP and compared the result with 18F-FDG. The picture is of feasibility demonstrated in small cohorts, plus one high-profile withdrawal.
Somatostatin is a naturally occurring cyclic peptide hormone of 14 or 28 amino acids, named for its inhibitory endocrine action: it suppresses the secretion of insulin, glucagon, and other hormones. That inhibitory function is conserved across vertebrates. In ex vivo pituitary cultures from an iguana, a chicken, and a rat, somatostatin blocked growth hormone-releasing hormone GHRH -stimulated growth hormone release in all three species, even though the stimulatory responses to other peptides varied widely between them PMID 38994081 .
Somatostatin signaling extends far beyond the hormone axis. A 2025 study of neuropathic pain in mice found that somatostatin-positive interneurons in the medial prefrontal cortex were the decisive cell population: increased mGluR5 activity in those interneurons suppressed pyramidal neuron firing and produced mechanical allodynia, and optogenetic silencing of the interneurons reversed the increased inhibitory input PMID 40211051 . The relevance to imaging is indirect but real. Somatostatin is made and sensed in many tissues, so the success of somatostatin-based imaging does not come from the peptide being rare. It comes from where the receptors are.
The five somatostatin receptor subtypes, SSTR1 through SSTR5, are G-protein-coupled receptors overexpressed in neuroendocrine tumors, including pancreatic neuroendocrine tumors, small cell lung cancer, and thyroid carcinoma. Synthetic analogs such as octreotide acetate, pasireotide, and vapreotide were developed to exploit these receptors, and radiolabeled versions of the analogs are used for imaging and therapy of SSTR-expressing tumors. Among the nine systems in this review, this is the only one whose imaging use has reached standard clinical practice.
Bombesin and gastrin-releasing peptides form a neuropeptide family with a broad physiological portfolio: exocrine and endocrine secretion, thermoregulation, sucrose regulation, and cell growth. Four receptor subtypes are recognized: the neuromedin B receptor, bombesin receptor subtype 3, the gastrin-releasing peptide receptor GRPR , and bombesin receptor subtype 4. Bombesin-like peptide receptors are overexpressed in breast cancer, ovarian cancer, and gastrointestinal stromal tumors, and GRPR is the explicit target of two registered prostate cancer studies discussed below.
The breadth of GRP biology is a caution for imaging. Two comparative studies show how far the peptide's reach goes. In the trigeminal sensory system, GRP is expressed in small-diameter ganglion neurons, about 6% of mouse and 8% of shrew neurons, with dense fibers in the superficial caudal trigeminal spinal nucleus, a pathway that mediates itch and is conserved across mammals PMID 28127106 . In the lumbosacral spinal cord, GRP forms a sexually dimorphic, androgen-dependent system with male-specific expression that controls reproductive function in both mice and Asian house musk shrews, again indicating conservation PMID 27804131 . A peptide active in itch and reproduction is expressed in normal sensory and spinal tissue, and any GRPR-targeted tracer must live with that background.
The receptor chemistry, meanwhile, is well characterized. In 1991, C-terminally modified GRP antagonists, in which the Leu26-Met27 region was replaced by an alkyl group, proved equipotent with native GRP at an IC50 of 2 nM, resisted enzymatic degradation, and blocked both GRP-stimulated fibroblast mitogenesis and GRP-induced calcium signaling in small cell lung cancer cells PMID 2066982 . The work is in vitro and decades old, but it established that the GRPR pocket tolerates substantial modification at the C-terminus, which is what makes radiolabeled GRP analogs feasible.
Two registered studies bring GRPR imaging into patients. NCT04746638, a phase 1 study, completed with 10 patients with prostate or breast cancer using a technetium-99m-labeled GRPR-targeting peptide, RM26. NCT05633160, a phase 1/2 study of copper-64 and copper-67 labeled bombesin analogs for imaging and treatment of GRPR-expressing metastatic castration-resistant prostate cancer in patients ineligible for 177Lu-PSMA-617 therapy, was terminated after enrolling 4 patients. The contrast is worth noting: the imaging study completed, the therapeutic companion did not.
VIP is a 28-amino-acid neuropeptide with an average molecular weight of 3326.78 and molecular formula C147H237N43O43S1; the single sulfur atom is the methionine in the sequence. VIP promotes vasodilation and stimulates cell growth through the VPAC1 and VPAC2 receptors, which are highly expressed in brain tumors, pancreatic adenocarcinomas, and neuroendocrine tumors. VIP has been investigated as a therapeutic agent, but toxicity may occur even at sub-microgram doses, a constraint that limits any systemic use and argues for restricting the peptide to tracer quantities. No registered VIP imaging trial was identified in the records examined for this article.
CCK and gastrin are structurally and functionally related peptides whose shared bioactive sequence is eight amino acids, with an average molecular weight of 1144.25 and molecular formula C49H61N9O17S3. The two hormones are identical in that bioactive sequence and differ in a single chemical detail: the site of tyrosine sulfation, position 6 in gastrin and position 7 in CCK. The formula carries three sulfur atoms, one of which is the sulfate ester on that tyrosine. The CCK2/gastrin receptor is frequently implicated in human cancers, including ovarian stromal tumors and astrocytomas. As with VIP, no CCK- or gastrin-based imaging trial appears in the registered studies examined here.
Neurotensin is a 13-amino-acid peptide with an average molecular weight of 1672.92 and molecular formula C78H121N21O20, according to the catalog description. The same catalog's data section lists the peptide at 12 amino acids. That internal discrepancy matters. A commercial listing that cannot agree with itself on the length of its own product is a concrete argument for confirming sequence and molecular weight by mass spectrometry before use. Neurotensin receptors have been identified in ductal pancreatic adenocarcinoma, small cell lung carcinoma, and medullary thyroid carcinoma, which has made the peptide a pursued candidate for cancer imaging. Reagents include the C-terminal fragment neurotensin 8-13 and its analogs.
α-MSH is a linear tridecapeptide of 13 amino acids, average molecular weight 1623.83, molecular formula C75H106N20O19S1, best known as the peptide that regulates skin pigmentation. It binds the melanocortin-1 receptor MC-1R with high affinity, and MC-1R is expressed in over 80% of human melanoma metastases. That combination, a defined receptor and a tumor type that almost always carries it, has made α-MSH a pursued carrier for targeted imaging and radiotherapy of melanoma. Because the same receptor controls pigmentation in normal melanocytes, the imaging window depends on receptor density in metastases exceeding that in normal skin. The reagent is commonly supplied as the C-terminally amidated peptide, α-MSH amide. No registered trial of α-MSH-based imaging appears in the records examined here.
T140 is a short, disulfide-constrained peptide: 14 amino acids, one disulfide bond, average molecular weight 2039.44, molecular formula C90H143N33O18S2. It acts as an antagonist of CXCR4, the chemokine receptor, meaning it occupies the receptor without activating it. T140 derivatives have been developed as imaging agents targeting CXCR4. The source material assigns no specific tumor indication to T140, and no T140-based trial appears in the registered studies examined.
Exendin-4 is a 39-amino-acid peptide hormone that shares 50% sequence homology with GLP-1 and acts as an agonist of the GLP-1 receptor. The source treatment of exendin-4 is incomplete, truncated mid-description, and the registered studies examined here contain no exendin-4 trial. Its clinical status for imaging or therapy is therefore unresolved in this record.
The evidence base across these nine systems is strikingly uneven. Somatostatin receptor imaging is mature clinical practice. RGD and GRPR systems have completed small human feasibility studies, with 16, 10, and 27 patients in the completed trials reviewed here, against one withdrawn study NCT01176500 and one terminated study NCT05633160 . The remaining systems, VIP, CCK/gastrin, α-MSH, neurotensin, and exendin-4, are represented in this record only by catalog entries and, in the case of α-MSH, by the claim that over 80% of melanoma metastases express the target receptor. Absence of a registration is not evidence that a system fails. It is evidence that the system has not yet generated publicly registered clinical data.
Several literature findings reinforce the need for caution. The GRP antagonists that looked potent in vitro in 1991 PMID 2066982 proved potent in vitro; whether they later produced a dominant clinical imaging agent is not addressed by that source. The conserved roles of GRP in itch PMID 28127106 and male reproductive function PMID 27804131 , and of somatostatin interneurons in cortical pain processing PMID 40211051 , all demonstrate that none of these receptors is tumor-exclusive. Imaging always operates against a background of normal receptor expression, and the quality of a scan is determined by the density difference, not by the mere presence of the receptor.
A researcher choosing among these systems should work through five checks. First, match the receptor to the tumor type using the expression data in the table above, and confirm receptor expression in the specific indication by immunohistochemistry, autoradiography, or prior imaging rather than assuming it from the catalog literature. Second, verify the reagent. The neurotensin length discrepancy is the textbook case: confirm sequence, molecular weight, and, for disulfide-containing peptides such as T140, correct oxidative folding. Third, match the radionuclide to the platform. The registered studies use 99mTc for SPECT and 18F, 68Ga, and 64Cu for PET, and the isotope half-life should be matched to the peptide's clearance kinetics so that signal is acquired while the tracer is still target-bound. Fourth, decide deliberately between an agonist and an antagonist at the target receptor. T140 exemplifies the antagonist class: it binds CXCR4 without activating the receptor, which avoids provoking the very signaling the tracer is meant to visualize. Fifth, budget for normal-tissue signal. A GRP-based tracer will encounter the peptide's territory in sensory ganglia and spinal cord, a somatostatin tracer will encounter cortical interneurons, and an α-MSH tracer will encounter melanocytes.
Four questions remain open on the evidence presented here. The clinical status of exendin-4 is unknown because the source text was truncated and no trial registration was located. The relative specificity and sensitivity of the nine systems for their tumor indications have not been compared head-to-head in any study supplied. The registration for NCT01176500 gives no reason for its withdrawal, and the supplied record for NCT05633160 does not indicate that its termination was due to the tracer itself; neither outcome can be attributed to the tracer itself. And the neurotensin length contradiction, 13 residues against 12, stands unresolved in the source materials, which is itself the strongest argument for treating catalog molecular data with skepticism until independently verified.
Peptides referenced: Exenatide, Growth Hormone, VIP (Vasoactive Intestinal Peptide), Octreotide, Pasireotide, Somatostatin, Glucagon, GLP-1.
Related reading: How cyclic peptides cross lipid membranes: a four-step mechanism, AI-Guided Peptide Library Design: Capabilities and Outcomes, Solid-Phase Peptide Synthesis: Resins and Working Protocols, Enzymatic Routes to Oligopeptide Synthesis: A Technical Overview.