Triple-Peptide Receptor Targeting In Vitro Allows Detection of All Tested Gut and Bronchial NETs

DISCUSSION

The in vitro comparison of single-receptor ligand binding using ¹²⁵I-Tyr³-octreotide, ¹²⁵I-GLP-1(7–36)amide, or ¹²⁵I-GIP(1–30) versus cocktail binding with all 3 radioligands in NETs shows several important features. First, it is evident, and as such confirms previous reports (2,6,7), that single ligand binding to each of the 3 receptors is good in a large number of NETs. However, there are with each single ligand a significant number of tumors that are not labeled or that have only a low density of receptors, a density that may not be sufficient to allow in vivo detection (11). Conversely, cocktail binding also reveals the expression of receptors, often in high density, for cases in which binding was negative with at least 1 of the single radioligands. On the basis of previous experience on the comparison of somatostatin receptor autoradiography with somatostatin receptor imaging, the high receptor density detected in the tumors of the present study with the cocktail suggests that virtually all these tumors should also become detectable in vivo with the cocktail (11).

Second, the cocktail may overcome the problem of heterogeneous receptor distribution seen with single radioligands, a heterogeneity that is occasionally observed in NETs and that may be a reason for insufficient imaging in vivo and for poorly successful targeted radiotherapy. For the success of targeted radiotherapy, it may indeed be decisive that the radioligand reaches every single cell of a tumor. The general problem of receptor heterogeneity and its consequence for adequate therapy has recently been emphasized (12).

To our knowledge, our study was the first to investigate such a triple receptor in vitro, although the idea for such a possibility had been introduced many years ago (2,13). However, methodologic considerations made it difficult up to now to have an in vitro cocktail experiment such as the present one performed convincingly. Indeed, to perform a cocktail experiment, it is crucial that the binding protocol be identical for each of the single radioligands, in order to have the cocktail experiment performed under controlled conditions suited to all 3 radioligands at the same time. Although the experimental protocol is different for somatostatin receptors, cholecystokinin receptors, and NPY receptors (2,13), the experimental protocol is, luckily, identical for somatostatin receptors, GLP-1 receptors, and GIP receptors (2,6,7), making it possible to investigate these 3 receptors simultaneously.

What are the consequences of these in vitro data for nuclear medicine? These data indicate that adequate radioligands for the detection of these 3 receptors in vivo have to be chosen: established radioligands exist for somatostatin receptor labeling (chelated octreotide and octreotate derivatives (14), or, more recently, chelated somatostatin receptor antagonist (15)) and for GLP-1 receptor labeling (chelated exendin derivatives (8)); radioligands for GIP receptors are in development, with the first in vivo animal imaging studies giving the proof of concept that GIP receptor targeting is feasible in vivo (16). Better radioligands need to be developed for therapy: specifically, current GLP-1 and GIP analogs have an excessive kidney uptake. Incretin receptor antagonists may be an option; they would also reduce the pharmacologic impact on glucose metabolism when 2 incretin receptor agonists are used concomitantly (8). Also, multivalent ligands may be an option, but it may be difficult to design multivalent ligands with a sufficiently high affinity for each of the 3 receptors. Then, the optimal radioligand dosage will have to be found and the complex logistic for using 3 different radioligands concomitantly has to be established. Since these ligands are labeled with radiometals, one could envision a procedure in which 3 peptide ligands are labeled in a single reaction vial, which could make multireceptor targeting clinically feasible.