Visual Abstract
Abstract
Our objective was to assess the value of 68Ga-DOTATOC and carbidopa-assisted 18F-fluorodihydroxyphenylalanine (18F-DOPA) in 21 hypoglycemic patients. Methods: All patients who underwent 68Ga-DOTATOC or carbidopa-assisted 18F-DOPA PET/CT for suspicion of insulinoma from January 2019 to January 2021 were retrospectively analyzed. A final diagnosis of insulinoma was determined by pathologic reports or consensus. Results: During the study period, 21 patients underwent both 68Ga-DOTATOC and 18F-DOPA PET/CT. A final diagnosis of insulin-secreting tumor was reached in 12 cases, including 11 insulinomas and 1 small mixed neuroendocrine/nonneuroendocrine neoplasm. 18F-DOPA and 68Ga-DOTATOC PET/CT were positive in 5 (45%) and 7 (64%) of 11 cases, respectively, with 4 concordant positive findings. Moreover, 1 insulinoma was visualized exclusively by 18F-DOPA PET/CT and 3 by 68Ga-DOTATOC PET/CT only. 18F-DOPA and 68Ga-DOTATOC PET/CT were falsely positive in 1 nonfunctioning pancreatic neuroendocrine tumor. Conclusion: When 68Ga-exendin-4 is not available, 68Ga-somatostatin receptor PET/CT should be the first choice for insulinoma functional imaging.
In adults, endogenous hyperinsulinemic hypoglycemia is commonly related to insulinoma. Insulinoma can induce severe debilitating and life-threatening hypoglycemia. The average incidence of insulinoma is 1–4 cases per million persons per year, and more than 90% of insulinomas are solitary, sporadic, and benign. In adults with hypoglycemia or suggestive symptoms, the diagnosis of hyperinsulinemic hypoglycemia relies on a positive fasting test (1). In 5%–10% of cases, insulinoma can occur in multiple endocrine neoplasia type 1. Parenchyma-sparing (enucleation/pancreatic resection) surgery is the optimal strategy (cure rate, 98%) but can be associated with complications. Accurate localization of insulinoma is therefore of primary importance.
Imaging work-up of patients with hyperinsulinemic hypoglycemia often requires a combination of anatomic and functional modalities (2). 68Ga-exendin-4, which targets the glucagonlike peptide 1 receptor, is expected to become the first-choice radiopharmaceutical for PET detection of benign insulinoma (3). However, it is available in only a few centers and is currently used only as a part of clinical trials. A special advantage of 68Ga-exendin-4 over other tracers is its unique value for distinguishing insulinoma from other neuroendocrine tumors—a condition of particular interest in the setting of multiple endocrine neoplasia type 1 patients, who often present with concomitant functioning and nonfunctioning pancreatic tumors. When 68Ga-exendin-4 is not available for clinical use, either 68Ga-radiolabeled somatostatin analog (68Ga-DOTA-SSA) or 18F-fluorodihydroxyphenylalanine (18F-DOPA) can be used (4). 68Ga-DOTA-SSA showed encouraging preliminary results for insulinoma imaging in 2 retrospective studies (5,6). However, insulinoma detection remains challenging, mainly because these tumors have a limited somatostatin receptor (SSTR) 2 expression profile and a small size and because of the physiologically high 68Ga-DOTA-SSA uptake in the uncinate process (7). On the other hand, the value of 18F-DOPA PET/CT is hampered by the relatively short duration of 18F-DOPA tumor retention in insulinoma and the diffuse uptake in normal pancreatic parenchyma, which may potentially mask insulinoma (8). To circumvent these major drawbacks, we previously proposed a revised imaging protocol based on a dual-phase imaging acquisition and patient premedication with carbidopa (a peripheral aromatic amino acid decarboxylase inhibitor) to prevent physiologic pancreatic 18F-DOPA uptake (9).
Currently, there remains a degree of uncertainty regarding the optimal first-choice tracer for patients in whom insulinoma is suspected in the absence of 68Ga-exendin-4. The aim of the present study was to describe the value of 68Ga-DOTA-SSA and carbidopa-assisted 18F-DOPA in a retrospective series of hypoglycemic patients evaluated by both tracers.
MATERIALS AND METHODS
Patients
This retrospective study was conducted in the departments of nuclear medicine of 3 university hospitals in France (Strasbourg, Marseille, and Nancy). We performed a comprehensive search of our databases to identify all patients evaluated by 68Ga-DOTA-SSA or carbidopa-assisted 18F-DOPA PET/CT for clinical, biologic, or radiologic suspicion of insulinoma-related hyperinsulinemic hypoglycemia from January 2019 to January 2021 (Fig. 1). Of the patients who underwent 18F-DOPA and 68Ga-DOTA-SSA PET/CT, only those studied within a 3-mo period without therapeutic intervention or a change in therapy between the 2 PET studies were included. All data (clinical, biologic, and imaging) were extracted from institutional medical data files (Table 1). The institutional review board approved this retrospective study, and the requirement to obtain informed consent was waived.
Flowchart summarizing study design and key PET/CT imaging findings. MiMEN = mixed neuroendocrine–nonneuroendocrine neoplasm; NF = not functioning.
Patients with Suspected Insulinoma Who Underwent Both Carbidopa-Assisted 18F-DOPA and 68Ga-DOTATOC PET/CT
Imaging Protocols
All examinations were performed by combined PET/CT devices equipped with 3-dimensional time-of-flight technology. The patients were injected with a 2–3 MBq/kg dose of 68Ga-DOTATOC and a 3–4 MBq/kg dose of 18F-DOPA (2 h after carbidopa premedication, 200 mg orally) without fasting beforehand. 68Ga-DOTATOC PET/CT included a whole-body acquisition from the upper thigh to the top of the skull (3–5 min/step), starting 60 min after radiotracer injection. Dual-time-point 18F-DOPA PET/CT included an early scan of the upper abdomen (at 5 min, 10 min/step) and a delayed whole-body acquisition (at 30 min, 3–5 min/step). In all cases, low-dose nonenhanced CT was performed and used for attenuation correction.
A pancreatic abnormality was defined as a focal area of increased uptake compared with surrounding tissue, considering potential pitfalls for both tracers. For 18F-DOPA PET/CT, a positive early phase followed by a negative delayed scan was considered a pathologic study. Semiquantitative analysis was performed by centering a spheric volume of interest on the uptake foci.
Gold Standard
The final diagnosis of insulinoma was determined by the pathologic results when available. In the remaining patients, the diagnosis was reached by a consensus considering clinical, biologic (positive fast test), and radiologic (CT/MRI typical enhancement) parameters and follow-up.
Statistical Analysis
The results for continuous data are expressed as the mean ± SD or the median and range as appropriate, whereas categoric variables are presented as numbers and percentages. Detection rate, sensitivity, and specificity are provided for both modalities.
RESULTS
Patient Population
In total, 32 patients were evaluated during the study period: 9 with 18F-DOPA PET/CT alone, 2 with only 68Ga-DOTATOC PET/CT, and 21 with both tracers (Fig. 1). The 21 patients imaged with both tracers constituted the study population. The patients’ characteristics are summarized in Table 1.
Fourteen of 21 patients had a positive 72-h fasting test, whereas 7 had doubtful results. In patients with borderline biochemical and imaging findings, the indication for further evaluation was decided in the setting of institutional multidisciplinary meetings.
A final diagnosis of an insulin-secreting tumor was reached in 12 patients, including 11 insulinomas (1 occult) and 1 small mixed neuroendocrine/nonneuroendocrine neoplasm with 15% insulin cell positivity. Among them, the fasting test was positive in 10 patients and inconclusive in the remaining 2 patients. Among the 11 insulinomas, 8 were pathologically proven; among the other 3 cases, the diagnosis was reached by consensus. In 7 patients without a detectable pancreatic target lesion on either anatomic and functional imaging, the diagnosis of insulinoma was excluded by follow-up. In the other 2 patients, the diagnosis turned out to be nonfunctioning pancreatic neuroendocrine tumors confirmed by surgery or endoscopic ultrasound-guided fine-needle aspiration biopsy and follow-up.
PET/CT Findings
The PET/CT findings are summarized in Figure 1. For insulinoma, 18F-DOPA and 68Ga-DOTATOC PET/CT were positive in 5 and 7 of the 11 patients, respectively, with concordant positive findings in 4 patients. Moreover, 1 insulinoma was visualized exclusively by 18F-DOPA PET/CT, and 3 were visualized only by 68Ga-DOTATOC PET/CT (Fig. 2). On 18F-DOPA PET/CT, delayed-phase imaging failed to detect 1 insulinoma that was correctly identified by early-phase 18F-DOPA and 68Ga-DOTATOC PET/CT (Fig. 2). In this patient, the insulinoma was related to multiple endocrine neoplasia type 1 syndrome, and the patient had a previous history of 3 benign insulinomas that were positive on somatostatin receptor scintigraphy and successfully treated by surgery. In the other patients, insulinomas were sporadic. One 10-mm grade 1 mixed neuroendocrine–nonneuroendocrine neoplasm remained occult on both 18F-DOPA and 68Ga-DOTATOC PET/CT, likely because of the small percentage of the neuroendocrine component.
Discordant results for carbidopa-assisted 18F-DOPA and 68Ga-DOTATOC PET/CT in 2 patients with pathologically proved insulinoma (arrows). (A) Early 18F-DOPA–positive/delayed 18F-DOPA–negative/68Ga-DOTATOC–positive findings. (B) Early 18F-DOPA–negative/delayed 18F-DOPA–negative/68Ga-DOTATOC–positive findings.
Among 2 patients with nonfunctioning pancreatic neuroendocrine tumors, 18F-DOPA and 68Ga-DOTATOC PET/CT were falsely positive in one and negative in the other. The diagnosis of insulinoma was excluded in 7 additional patients without identifiable pancreatic target images on either PET imaging study. The overall detection rate, sensitivity, and specificity were, respectively, 45%, 42%, 89% for 18F-DOPA and 64%, 58%, 89% for 68Ga-DOTATOC PET/CT.
The tumor-to-background uptake ratios were more favorable with 68Ga-DOTATOC than with 18F-DOPA PET. The mean ratios of tumor SUVmax and tumor SUVmax to normal pancreas SUVmean were 6.7 and 1.5, respectively, for early-phase 18F-DOPA; 6.3 and 2.7, respectively, for delayed-phase 18F-DOPA; and 45.2 and 10.8, respectively, for 68Ga-DOTATOC PET/CT.
DISCUSSION
The present study aimed to describe the value of 68Ga-SSTR PET/CT and carbidopa-assisted 18F-DOPA in a series of hypoglycemic patients. The principal conclusions were, first, that 68Ga-DOTATOC PET/CT has a high detection rate in insulinoma, although its value is less than that in nonfunctioning pancreatic neuroendocrine tumors because of a lack of SSTR2 expression in a subgroup of insulinomas (10), and, second, that 68Ga-DOTATOC PET/CT can be positive when 18F-DOPA fails (3 cases) and vice versa (1 case). The latter point is easily comprehensible because of the various molecular determinants of tracer uptake and retention in both conditions.
Although our study was not designed to perform a reliable comparison between 68Ga-SSTR PET/CT and 18F-DOPA (mainly because the study was retrospective and had a limited number of patients), we suggest using 68Ga-SSTR PET as the first-choice tracer when 68Ga-exendin-4 is not available. This position could be supported by several arguments. Two recent retrospective studies showed promising results concerning the use of 68Ga-SSTR PET/CT in patients with insulinoma-related hyperinsulinemic hypoglycemia, allowing for the identification of pancreatic secreting tumors in 9 of 10 (90%) and 11 of 13 patients (85%) (5,6). The greater sensitivities described in these studies than in our study could be related to the inclusion of solely pathologically proven insulinomas (5,6), excluding patients with nonoperated 68Ga-SSTR–negative insulinoma. Furthermore, on the basis of the widely admitted expression of SSTR2 in two thirds of insulinomas (10), it is expected that 68Ga-SSTR PET/CT sensitivity in real-life situations should be less than previously reported, with an on–off uptake pattern, depending on SSTR2 expression. The use of 68Ga-SSTR PET/CT also has practical advantages over 18F-DOPA in terms of availability and cost for teams skilled and suitably equipped for 68Ga radiolabeling.
Although the application of carbidopa-assisted 18F-DOPA PET/CT remains controversial (11), the present study showed that it can be useful for SSTR-negative insulinoma and therefore can be considered a second-choice tracer when 68Ga-SSTR PET/CT fails to detect the tumor. The lower rate of 18F-DOPA PET/CT positivity than previously reported could also be related to selection bias. As highlighted in the flowchart, 5 insulinomas, including 4 with 18F-DOPA–positive findings, did not undergo 68Ga-SSTR PET/CT. With these patients, the rate of positivity would therefore be 56% (9/16). Physicians should be aware that imaging protocols should be adapted, including carbidopa premedication. We previously showed in a preclinical model that the use of carbidopa did not inhibit insulinoma 18F-DOPA uptake (12), a phenomenon that was described for β-cell hyperplasia (13). In the present study, delayed acquisition missed 1 insulinoma, and early acquisition was never inferior to delayed acquisition. In a previous study that included 24 patients, 4 cases were detected only by early-phase acquisition (14).
CONCLUSION
Despite the limitations that have been pointed out, this study provided new data on both tracers in this rare but curable disease. When 68Ga-exendin-4 is not available, we suggest using SSTR analogs as the first-choice PET tracer and considering carbidopa-assisted 18F-DOPA as a valid alternative when the results are inconclusive.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: What is the best radiopharmaceutical for insulinoma localization in the absence of 68Ga-exendin-4?
PERTINENT FINDINGS: When 68Ga-exendin-4 is not available, 68Ga-SSTR should be considered the first-choice PET tracer. Carbidopa-assisted 18F-DOPA PET remains a valid option when the results are inconclusive.
IMPLICATIONS FOR PATIENT CARE: 68Ga-SSTR PET/CT enables detection of insulinoma, allowing curative sparing surgery (enucleation/pancreatic resection) and resolution of preoperative symptoms.
Footnotes
Published online July 16, 2021.
- © 2022 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication April 3, 2021.
- Revision received May 27, 2021.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.