[⁶⁸Ga] labelled Exendin for radioguided surgery of intrapancreatic insulin producing lesions in patients with congenital hyperinsulinism

Abstract

Background

Congenital hyperinsulinism (CHI) is a life threatening disease. Localization of affected intrapancreatic beta cells responsible for focal forms during surgery can be challenging. In this study we investigated a new radioguided surgical (RGS) approach using [⁶⁸Ga]Exendin to facilitate intraoperative focus detection. All patients were scanned initially with [¹⁸F]-DOPA-PET followed by [⁶⁸Ga]Exendin PET to differentiate between focal and non-focal forms. Focal CHI patients were then operated. At the beginning of standard surgical dissection of the pancreas in CHI patients (n = 12), 46 MBq of [⁶⁸Ga]Exendin were injected intravenously. Intrapancreatic localization of the foci was determined by using a hand-held positron- and gamma-radiation probe. RGS was carried out as enucleation of CHI foci. Duration of surgery (defined as the time lapse from first incision until final suture placement) for RGS was compared with historical data of patients operated on without RGS. Long term follow-up data on euglycemic control were retrieved from patient´s medical files.

Results

[¹⁸F]-DOPA- and [⁶⁸Ga]Exendin PET findings were concordant in all patients. Overall, 12 CHI patients underwent RGS. In 11/12 children (92%) the CHI foci localized pre-operatively by [⁶⁸Ga]Exendin PET could be detected intraoperatively using the hand-held positron probe. There was a high correlation between PET imaging results and positron probe findings in respect to the identification of the affected pancreatic region. One pancreatic lesion in close proximity to the left kidney could not be detected by the positron probe. Histopathology confirmed all resected lesions as CHI foci. Intraoperatively, the signal of the focus was > 10 times higher than the signal of normal adjacent pancreatic tissue. Median duration of surgery was 4.7 h (CI 3.5–6.7) in RGS patients compared to 5.5 h (CI 4-6.7) in patients undergoing surgery without radio guidance. All patients remained euglycemic after surgery (median follow-up 3 years, range 2 to 4.5).

Conclusions

In this study, we demonstrated the use of [⁶⁸Ga]Exendin for intraoperative localization of intrapancreatic CHI foci. RGS facilitates localization of intrapancreatic CHI focus and thus potentially reduces duration of surgery and perioperative complications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13550-025-01294-8.

Background

Congenital hyperinsulinism (CHI) is a rare disease characterized by presence of functionally altered non-neoplastic beta-cells with inappropriate (over-) secretion of insulin, leading to life-threatening hypoglycemia and thereby to an increased risk for brain injury and cognitive impairment. The disease can either be focal or diffuse, indicating that either one circumscribed pancreatic region or all pancreatic beta cells are affected. In addition atypical CHI forms have in been some cases described. The increased insulin secretion in focal CHI forms is in the majority of cases related to a genetic defect on the paternal allele in one of the potassium channel subunits (potassium inwardly rectifying channel subfamily J member 11, KCNJ11; ATP binding cassette subfamily C member 8, ABCC8) accompanied by a cell-specific loss of heterozygosity on the maternal allele. Patients with focal CHI forms can be cured by resection of the affected region, whereas diffuse CHI forms are mainly treated pharmacologically. The benefit of surgery in this latter group is matter of a controversial discussion.

Curative surgical treatment is the primary aim in patients with focal CHI forms. Preoperative [¹⁸F]DOPA PET imaging (¹⁸Ffluoro-L-3,4dihydroxyphenylalanine positron emission tomography) in combination with 3-phase contrast CT or MRI has been established as the state-of-the-art diagnostic procedere to localize the intrapancreatic focus in CHI [1]. If a focus is identified, surgical resection is performed by enucleation or partial pancreatectomy, depending on the location of the focus. Small size of the pancreas in infants, artefacts, insufficient or missing tracer uptake [2–4] and sometimes challenging focus localizations near the pancreatic duct, the bile duct or large vessels could make surgery very difficult and increase the risk for complications or incomplete focus resection. Even with information from PET imaging the intraoperative localization and definition of focus-borders is still challenging because it is neither visible nor palpable in at least one third of cases [5]. Standards of surgical care involve repeated intraoperative biopsies for confirmation of the disease type and later on for assessing complete resection of the focus by frozen section histopathology. This approach, although a worldwide standard, is complicated as the surgeon has to find an optimal compromise between resecting all affected tissue to prevent persistent hypoglycaemia and at the same time avoid resecting too much unaffected pancreatic tissue. Resection of too much tissue carries the risk for diabetes mellitus, exocrine pancreatic insufficiency and surgical complications, which may result in reduced quality of life or even overall survival [1, 6, 7]. In focal CHI forms a pre-operative diagnostic accuracy of 74.5% [8] has been described using [¹⁸F]DOPA PET imaging. However, worldwide availability of [¹⁸F]DOPA especially in middle-income countries is limited due to the need of a medical cyclotron for tracer production. Furthermore, false negative results have been observed.

Exendin-4 peptide binds to the glucagon like peptide 1 receptors on pancreatic beta cells with high degree of affinity and also gets internalized [9]. In several prior studies, [⁶⁸Ga]exendin has been found to have excellent diagnostic performance for detection of insulinomas [10, 11]. In a previous multi-center study the diagnostic performance of novel tracer [⁶⁸Ga]NODAGA Exendin 4 (Exendin), in comparison to [¹⁸F]DOPA, has been evaluated for detection of focal CHI lesions [12]. It was observed that [⁶⁸Ga]NODAGA Exendin 4 has a higher sensitivity and better interobserver correlation than [¹⁸F]DOPA.

Image-guided surgeries in small pediatric cohorts have their own challenges. In addition to improving differentiation between focal and diffuse CHI forms pre-operatively, in-vivo visualization of the focus during surgery using radiolabelled probes would be a significant improvement for identification of the affected region and overall outcome of this treatment.

Hand-held probes for radioguided surgery (RGS) have been used successfully in surgery of sentinel lymph nodes, parathyroid gland, in breast and colorectal cancer and insulinoma [13–15]. However prior probes for RGS studies used gamma radiation for localization. One of the limitations of gamma probe for RGS is lower spatial resolution. In comparison, probes that detect positrons offer higher spatial resolution [16]. This prompted us to postulate that intraoperative radio-guided surgery using specific radiopharmaceuticals would be of significant help to detect intrapancreatic CHI foci and navigate the surgeon to the affected pancreatic region. To localise the lesion precisely, radiopharmaceuticals should show a high specificity, a high target to non-target ratio, preferably a very low uptake in normal pancreas, and should be readily available to be adjusted in the complex time schedule of the operation. In contrast to [¹⁸F]DOPA, the new tracer Exendin can be prepared in-house within an hour whenever required. Furthermore, [¹⁸F]DOPA is not optimal for RGS of the pancreas due to tracer accumulation in the liver and gall bladder, which reduces the capacity for focus detection.

With this background, in this explorative single cenetr study we explored the feasibility and usefulness of RGS in patients with focal CHI. We also assessed the specificity and safety of the RGS. In addition we compared the duration of surgery, defined as the duration from first incision until final suture placemen, to data of a historical age matched control group undergoing standard non-radioguided surgery at our institution.

Methods

Cohorts

All parents were informed in detail about the procedures including [⁶⁸Ga]Exendin PET and the repeated injection of the tracer intraoperatively for guiding the hand-held probe for foci detection. Imaging and application of the radiotracer intraoperatively was performed after parents gave written informed consent. Institutional ethical committee review (EA No. EA2/150/18) approved the analysis of the data. Compassionate usage of [⁶⁸Ga]Exendin in accordance with national drug regulations was applied because individual advantages were expected by reducing the risk of associated surgery and prolonged anesthesia.

This project is based on a previous study, in which individuals with CHI received both, a [¹⁸F]DOPA PET- and an [⁶⁸Ga]Exendin PET-scan [12]. All included patients received both − [¹⁸F]DOPA PET-low dose CT and [⁶⁸Ga]Exendin PET-MRI as imaging diagnostics (Fig. 1). Only focal CHI patients were referred for referred for RGS.

Fig. 1.

Study workflow: Patients received [¹⁸F]DOPA PET CT and [⁶⁸Ga]Exendin CT/ MRI. In case of a focal tracer accumulation, [⁶⁸Ga]Exendin was injected during surgery. CHI focus was identified using a hand-held cable-free positron probe. After operation, extracted pancreatic tissue was histologically analysed

Twelve children (male: n = 10; female: n = 2) with focal CHI (Table 1) were operated on using RGS. In all CHI patients hypoglycemia had occured directly after birth. CHI was diagnosed based on elevated fasting insulin levels without detectable levels of ketone bodies according to national guidelines. Additionally, genetic testing for ABCC8 and KCNJ11 was performed in all patients.

Table 1.

Details of the clinical characteristics of the two cohorts treated with or without radioguided surgery

As a control group we retrospectively analyzed the results of 10 CHI patients (male: n = 6; female: n = 4), who had been operated without RGS. For two patients within the non-RGS control group, who had been referred by peripheral centers, the exact annotation of the underlying genetic variants is missing.

All patients were initially examined with [¹⁸F]DOPA PET/CT (low dose CT) or [¹⁸F]DOPA PET/MRI in order to confirm either a focal or diffuse nature of the disease and to localize a potential focus. This was followed by a second PET/MRI with [⁶⁸Ga]Exendin as published recently [12].

Radiolabeling of ⁶⁸Ga Exendin

Radiolabeling of [⁶⁸Ga]Exendin was performed locally at our institutional radiochemistry laboratory (Charité - Universitätsmedizin Berlin). The protocol for labeling and quality control tests has already been described in detail previously [12, 17].

Imaging protocols for [¹⁸F]DOPA PET and [⁶⁸Ga]Exendin PET

[¹⁸F]DOPA PET/CT was performed according to standard protocols [1]. [⁶⁸Ga]Exendin PET imaging (low dose CT or MRI) was performed 30 min after injection of a target dose activity of 46 MBq [⁶⁸Ga]Exendin. To avoid radiation exposure from low dose CT, patients were imaged only from liver to pelvis. The results have been previously described [12].

All PET/CT scans were acquired in a 3-dimensional acquisition mode on a Gemini TF 16 PET/CT system (Philips Medical Systems). The standard 3D-LOR algorithm of the system software was used with default parameter settings to reconstruct transaxial slices of 144 × 144 voxels with 4.0 × 4.0 × 4.0 mm³. In all patients, a low-dose CT acquired immediately before the PET scan was used for attenuation correction (120 kVp, 30 mAs). Images were analyzed independently, unblinded, by two of the three nuclear medicine physicians (VP, WB, CF).

After administration of 46.0 (range, 25.0–86.0) MBq [⁶⁸Ga]Exendin, PET/MRI examination of the upper abdomen (Biograph mMR, Siemens Healthineers, Erlangen, Germany) was performed as a simultaneous measurement of PET and a 3-Tesla MRI in a combined scanner. The 3D PET measurement was performed for a total of 30 min, similar to the measurement period for the PET/CT examination with [¹⁸F]DOPA. While the first 20 min after tracer administration in PET were acquired dynamically, PET data from 20th to 30th minutes were reconstructed statically (ordered subset expectation maximization [OSEM], 3 iterations, 21 subgroups, Gaussian filter, 4 mm full width at half maximum, matrix of 172 × 172 × 127, voxel size of 4.17 × 4.17 × 2.03 mm³).

The MRI part of the simultaneous examination of the upper abdomen included the following sequences: fast 3D T1-weighted (Controlled Aliasing in Parallel Imaging Results in Higher Acceleration [CAIPI HiRes]) with standard settings (resolution, 1.3 × 1.3 × 3mm³; matrix, 312 × 384 × 88) for attenuation correction of PET data, a fast spin echo sequence (Half Acquisition Single-Shot Turbo Spin Echo [HASTE], resolution, 0.812 × 0.812 × 4mm³; matrix, 320 × 320 × 50), and a contrast-enhanced T1-weighted sequence (Volume Interpolated Gradient Echo [StarVIBE], resolution, 0.94 × 0.94 × 1.1mm³; matrix, 288 × 288 × 96; gadobutrol (9 patients), median, 1.0 (range, 0.5–1.5) ml; gadoterate meglumine (1 patient), 3.0 ml).

All examinations were performed under general sedation of the patients according to the institutional guidelines.

The lesions were considered positive if the uptake was focal, circumscribed and higher than the surrounding pancreas. Because the lesions in general are smaller than 5–10 mm resulting in substantial partial volume effects we did not use the SUV (standardized uptake value) for quantifying the uptake. Lesions were described visually on a three-point scale: mild: uptake just above the normal pancreas, moderate: uptake definitely more than the normal pancreas tissue around the focus, intense: very high uptake.

Positron probe

During surgery the PET probe system of First Sensor AG (Berlin, Germany) was used. The spatial resolution (full width at half maximum, FWHM) for ⁶⁸Ga gamma rays (511 keV) is approx. 13 mm and for positrons it is approx. 7 mm. With the gamma sensor it is possible to find the correct direction to the target, and with the positron detector it is possible to find the precise point of the target within the tissue.

Intraoperatively, the probes were first wrapped in a sterile pack and then moved from the normal pancreas towards the suspected focus (Fig. 1). A nuclear medicine physician (VP, WB, CF) was present during the operation to confirm if the counts detected by the probe were specific for a focus or not. On an average, 30 min after injection of the radiotracer, the probe was placed over the pancreas and surrounding organs to intraoperatively localize the lesion by radiation signal detection. A ratio between positron to gamma count close to one and/or the positron count ten times higher than the normal pancreas were considered indicative for a focal lesion.

For assessing the radiation exposure to the personnel in the theatre and pathology, we calculated both the hand doses and the whole-body doses for the various personnel based on mean injected activity and mean duration of the surgical procedure.

Procedures during and after surgery

After identification of the aspired lesion with the probe, enucleation was performed as atypical pancreas resection of the focus. After resection of each lesion, the probe was used to evaluate the resection surface on the pancreas for complete removal of the radioactivity which was assumed when the probe count rate decreased to background levels. Duration of surgery (defined as time lapse from first incision until the final suture placement) for RGS was compared with data of a historical control group consisting of patients operated on without RGS at our institution. Specificity and accuracy of RGS for detection of CHI focus was evaluated by comparing the results with histopathology and PET/CT (PET/MRI) results, respectively. Long term (3 years) follow-up data on euglycemic control were retrieved from patient´s medical files.

Results

Patient cohort

The clinical details of each patient are summarized in Table 1. In a total of twelve patients (male: n = 10; female: n = 2), radioguided surgery has been performed during operation. The median age at surgery was 0.4 years (CI 0.26 to 0.68). In 85% of the patients, a heterozygous mutation in the ABCC8 gene had been identified. The control group consisted of 10 patients with focal CHI forms related to 12 operations (male: n = 6; female: n = 4). Medium age at operation was 0.32 years (CI 0.17 to 0.52).

[⁶⁸Ga]Exendin PET radioguided surgery

The time interval between [¹⁸F]DOPA and [⁶⁸Ga]Exendin PET as well as between PET imaging and surgery ranged between 1 and 7 days. Eleven out of the twelve PET positive intrapancreatic lesions could be detected using the hand-held positron probe (Fig. 1). One lesion localized in the near vicinity of the kidney could not be localized with the hand-held positron probe because no localized lesion above background could be detected. This was due to overlapping PET tracer background signal, as the tracer is excreted via the kidneys. The size of the lesions detected by radioguided surgery ranged from 3 to 10 mm. Lesions were successfully detected within 30 min to 1 h after injection of the radiopharmaceutical in all patients (Fig. 2). Histopathology confirmed all lesions to be a CHI focus. Intraoperatively, the signal of the focus was > 10 times higher compared to unaffected neighboring tissue. None of the patients had developed any late complications by the time of discharge. All patients were free of hypoglycemic episodes during the postoperative stay in the hospital and were discharged in euglycemic state. Median duration of surgery was 4.7 h (CI 3.5 to 6.7) in children undergoing RGS as compared to 5.5 h (CI 4 to 6.7, p > 0.05) in the control group (Fig. 3).

Fig. 2.

Example of [¹⁸F]DOPA PET/CT and [⁶⁸Ga]Exendin PET/MRI images of a focal CHI patient. Upper left panel are [¹⁸F]DOPA PET axial slice (upper panel) and [¹⁸F]DOPA PET/MRI fused images (left lower panel) of a patient with CHI focus in the head of the pancreas (black arrow). The right upper panel is showing [⁶⁸Ga]Exendin PET axial slice whereas the right lower panel is showing [⁶⁸Ga]Exendin PET/MRI fused image

Fig. 3.

A Median duration of surgery in patients with radioguided surgery and control group. B Mean heart rate and mean arterial blood pressure (C) after tracer injection. Time points represent minutes. hrs = hours, * p < 0.05 (Mann Whitney test)

Intraoperative monitoring of tolerance of [⁶⁸Ga]Exendin and long-term outcome

Vital parameters were measured during surgery based on standard-operating procedures. After tracer injection, a significant increase of heart rate from median 124 beats per minutes (CI 115 to 135) at 20 min before tracer injection to 142 (CI 127 to 149.5) after 1 min after tracer injection (p = 0.04) was observed in all patients (Fig. 3). No further complications were identified during surgery. Two patients developed a pancreatic pseudocyst, and one a chyle-leak. However, these three patients fully recovered. All patients remained euglycemic after surgery with a median follow-up of 3 years (range 2 to 4.5). In the control group, one sepsis occurred and in two cases hypoglycemic episodes re-occurred.

For a mean duration of the surgical procedure of 4.7 h and a mean intravenously injected activity of 46 MBq [⁶⁸Ga]Exendin, the following doses would be achieved: hand-related dose for the surgeon (maximum: 10 cm distance during the whole procedure) 1.1 mSV, whole-body dose (40 cm distance) 0.07 mSv, further staff members (e.g. nurses, anesthetists; 180 cm distance) 0.003 mSv. For the pathologist, considering 10% of the whole injected activity being taken up by the pancreas and resected in the pathology specimens, and a handling time of 30 min, the hand-dose (10 cm) would amount to 0.02 mSv, the whole-body dose to 0.001 mSv. With respect to an annual limit of 1 mSv whole-body dose and 50 mSv hand dose to non-radiation exposed workers in Germany, even the surgeon being the most exposed individual would be allowed to perform up to 14 procedures per year under these maximal exposure scenario conditions without any need for radiation safety measures.

Both surgeons agreed, that RGS was useful in guiding them to the intrapancreatic CHI focus and expressed, that they and their team would prefer to perform radio-guided surgery instead of the classical surgical non-radio-guided approach (data in supplemental material).

Discussion

Surgery of focal CHI is challenging due to several factors. During operation, patients may have a different position on the operating table compared to the situation during imaging, making it difficult to use the images of pre-operative PET/CT or PET/MRI for precise localization of a usually non-visible small focus intraoperatively. In addition, the exact position of the pancreas itself can vary due to gastric/enteral filling with air during intubation. The current gold standard for localizing CHI focal lesions is [¹⁸F]DOPA PET, which can visualize beta-cell activity by the tracer being trapped in secretory granules giving only indirect information about the affected beta-cell mass⁴. All these technical limitations impair the reliability of intraoperative lesion detection, making it compulsory to confirm the suspected focus by repeated frozen-section biopsies which prolong the surgical procedure significantly. In this single-center approach in 12 children with suspected focal CHI forms we confirmed previous results, in which a high grade of correlation has been shown between the visualized pancreatic region in focal CHI forms by [¹⁸F]-DOPA-and [⁶⁸Ga]Exendin PET [12]. We could show that readily available [⁶⁸Ga]Exendin is an excellent tracer for intraoperative focus detection with a comparable specificity and sensitivity [12]. For this reason we took the next step and used this observation to evaluate, whether this tracer could be used to navigate the surgeon during operation as it has been shown for the operation of different other tumor types [18, 19].

In all RGS patients except one, the CHI focus could be detected with the hand-held probe. Due to renal excretion and tubular reabsorption, exendin has a very high kidney uptake resulting in high background activity. This explains why the hand-held probe could not detect an intrapancreatic CHI focus located close to the kidney in one patient. It is difficult to compare effectiveness of surgery in patients with CHI, because severity, focus size and individual localization within the pancreas of the focus are major confounding factors. In our study, the CHI focus was identifed in the majority of cases of the RGS group within the pancreatic head (83,3% versus 50% in control group), which impedes the surgical resection of the affected region due to its localization near the pancreatic duct and large vessels. We used the duration of surgery as one outcome parameter. This factor has its own limitation, being directly dependent on anesthesia, lesion localization as well as the duration taken by the pathologists to evaluate and confirm the CHI focus on frozen section biopsies. Despite these limitaions we observed that the median duration of surgery was reduced using RGS. Follow-up studies with larger samples sizes are needed to evaluate, if and to which extent radioguided surgery could be superior to the standard surgical procedure. Here, we describe for the first time the successful use of radioguided surgery in CHI patients in general.

However, there is a need to improve this advanced technology. The current hand-held probe is not flexible, the resolution is– in relation to possible focus-sizes– still too low and its success depends primarily on the position of the probe in relation to the surface of the pancreas and the anatomical relation to nearby normal organs with physiological tracer uptake like the kidney. From a surgeons’ point of view, a flexible probe would be easier to handle and possibly even allow localization of foci in vicinity to the kidney. Resection of a lesion detected by a positron probe is only a small success if the lesion is not completely removed. Because the positron probe has a resolution of only 7 mm, it cannot be used to reliably confirm complete resection and therefore it is still necessary to take further biopsies, even from potentially unaffected tissue, to confirm complete removal by histopathology. Sometimes, this is in itself a challenge because focal lesions are often only a few millimeters in size and fresh frozen specimens are not always conclusive. Thus, improving resolution of hand-held probes would foster and improve radioguided surgery in CHI.

Besides the main outcomes we observed a significant increase in the heart rate but no change of the mean blood pressure shortly after tracer injection. Dose dependent cardiac effects of Exendin have been reported in animals [20, 21] but data in human subjects are rare and for the first time reported in children here. This observation should be considered as potential risk for cardially endangered patients and incorporated into further clinical studies. Finally, although not analysed systematically, we observed a tendency towards lower blood glucose levels after tracer injection which might be related to the activation of GLP-1 receptors. This indicates the importance of careful blood glucose monitoring during the use of [⁶⁸Ga]Exendin as a tracer. Finally, it needs to be considered that the current state of the art tracer [¹⁸F]DOPA is not available in several countries. Due to its non-cyclotron, generator-dependent in-house production, [⁶⁸Ga]Exendin might be a valuable alternative and of relevance for affected patients, especially in low- and middle income countries.

The main limitation of our study is the small cohort size, which is due to the fact that focal CHI is a very rare disease of less than 1 in 50,000 live births. This rareness is complicated by a broad phenotypic spectrum and varying severity of the disease. Additionally, the surgical challenges are dependent on focus localization and size. All these issues are leading to difficulties to define comprehensive outcome criteria comparing patients with and without radioguided surgery.

Conclusion

In this study, we demonstrated the use of ⁶⁸Ga Exendin for intraoperative localization of intrapancreatic CHI foci. RGS facilitates localization of intrapancreatic CHI focus and thus potentially reduces duration of surgery and perioperative complications.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analyses were performed by PK, VP, KH, WB, OB, LL. The first draft of the manuscript was written by PK and VP and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. PAT helped in the dosimetry aspect of the manuscript.

Funding

The research leading to these results has received funding from the European Community’s Seventh Framework Programme (FP7/2007–2013) under grant agreement no. 602812 (BetaCure) and HORIZON-HLTH-2022 project LightCure (no. 101080327). The authors gratefully acknowledge funding from the German Research Foundation (DFG, INST 335/543-1 FUGG) for PET/MR use. P.K. is supported by DFG (KU2673/6 − 1; KU2673/7 − 1).

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval

Institutional ethical committee review (EA No. EA2/150/18) approved the analysis of the data. This study was performed in line with the principles of the Declaration of Helsinki.

Informed consent

This is retrospective data analyses. Imaging and application of the radiotracer intraoperatively was performed after the parents gave written informed consent.

Consent to publish

The authors affirm that human research participants provided informed consent for publication of the images in Figure(s) 1 and 2.

Competing interests

MG is an inventor on, and holder of, the patent “Invention Affecting GLP-1 and Exendin” (Philipps-University at Marburg, June 17, 2009). Other authors disclose no other conflict of interest.