Systematic screening of imaging biomarkers for the Islets of Langerhans, among clinically available positron emission tomography tracers

Abstract

Introduction

Functional imaging could be utilized for visualizing pancreatic islets of Langerhans. Therefore, we present a stepwise algorithm for screening of clinically available positron emission tomography (PET) tracers for their use in imaging of the neuroendocrine pancreas in the context of diabetes.

Methods

A stepwise procedure was developed for screening potential islet imaging agents. Suitable PET-tracer candidates were identified by their molecular mechanism of targeting. Clinical abdominal examinations were retrospectively analyzed for pancreatic uptake and retention. The target protein localization in the pancreas was assessed in silico by –omics approaches and the in vitro by binding assays to human pancreatic tissue.

Results

Six putative candidates were identified and screened by using the stepwise procedure. Among the tested PET tracers, only [¹¹C]5-Hydroxy-tryptophan passed all steps. The remaining identified candidates were falsified as candidates and discarded following in silico and in vitro screening.

Conclusions

Of the six clinically available PET tracers identified, [¹¹C]5-HTP was found to be a promising candidate for beta cell imaging, based on intensity of in vivo pancreatic uptake in humans, and islet specificity as assessed on human pancreatic cell preparations. The flow scheme described herein constitutes a methodology for evaluating putative islet imaging biomarkers among clinically available PET tracers.

Introduction

Why is functional imaging of the neuroendocrine pancreas of interest? The answer lies in the unique anatomic function of the pancreas, which is involved in the secretion of a host of vital hormones. The majority of the endocrine tissue in pancreas consists of the Islets of Langerhans. These microscopic structures represent 1–2% of the pancreatic volume, are spread across the pancreas and have utmost importance for regulation of the blood glucose homeostasis.

Insulin is produced exclusively in the beta cells, which comprise approximately 60% of the human islets of Langerhans [1], and the dysfunction or destruction of these cells is implicated in the pathogenesis of type 1 and type 2 diabetes (T1D/T2D) [2], [3].

Importantly, among the total population of beta cells there is stratification in activity, function and parameters such as oxygenation (e.g. the existence of “dormant” beta cells) between different subgroups [4]. This implies that plasma measurements of beta cell function, (for example the response in release of c-peptide in response to glucose) will not correlate to the actual beta cell mass [5]. Thus, we cannot measure beta cell mass in vivo by current plasma biomarkers.

This state of affairs imposes a bottleneck in development of novel anti-diabetic therapies, aimed at beta cell regeneration, beta cell replacement and interventions to inhibit ongoing beta cell destruction. In addition, it severely limits our understanding of the relation between beta cell mass and the basic etiology of T1D and T2D, which is currently based upon post mortem examinations on subject in different stages of the disease.

In vivo molecular imaging techniques including positron emission tomography (PET) enable visualization and quantification of biological processes, by measuring the biodistribution, pharmacokinetics and target binding of radioactively labeled molecules (PET tracers/ligands). The potential role of a PET tracer in directly measuring beta cell mass is related to several important factors, including the density of target receptors in the beta cells compared to in exocrine cells, its target specificity and the delivery to the target tissue. The low proportion of beta cells in pancreas impose high requirements for tracer specificity and target receptor density, in order to avoid disturbance of the beta cells signal by the signal of the exocrine pancreas. Other medical imaging techniques have been utilized for this purpose with encouraging results such as magnetic resonance imaging, despite lower sensitivity [6].

A number of positron emission tomography (PET) tracers are routinely used for the diagnosis of neuroendocrine neoplasms (NENs) in clinical practice. In addition, there are a plethora of targets in the CNS for which selective PET ligands have been developed and tested in the clinical setting. Some of these tracers are anecdotally known to exhibit high pancreatic uptake, which could reasonably be derived from a selective targeting of the neuroendocrine pancreas. Therefore, these tracers should be assessed further for their suitability as imaging biomarkers for the endocrine pancreas. A case in point is the Vesicular Monoamine Transporter 2 (VMAT2) marker [¹¹C]Dihydrotetrabenazine, (DTBZ) which was developed for imaging of the CNS, but later proposed as a PET marker also for the Islets for Langerhans [7].

There is therefore a need for screening clinically available PET tracers for the possibility of imaging of the neuroendocrine pancreas given the low cost/large benefit associated with this endeavor.

We present a stepwise algorithm consisting of in silico, in vitro and retrospective in vivo assessment of the utility of different PET tracers in imaging of the neuroendocrine pancreas.

The algorithm is designed to be associated with low cost and minimal amount of lab work in mind, in order to facilitate rapid screening of several candidate tracers.

The algorithm is here used to evaluate four PET tracers with anecdotally high uptake in the pancreas; [¹¹C]Hydroxyephedrine ([¹¹C]HED) [8], [9], [⁶⁸Ga]DOTATOC [10], [¹¹C]Harmine ([¹¹C]HAR) [11] and [¹¹C]5-Hydroxytryptophan ([¹¹C]5-HTP) [12]. [¹¹C]Acetate ([¹¹C]ACE) and [¹⁸ F]Fluoro-deoxyglucose ([¹⁸F]FDG) were included as reference tracers (or negative controls for the algorithm), known a priori for non-islet selectivity or preference for uptake in exocrine tissue.