Diagnosis and Treatment Course of Insulinoma Presenting as Hypoglycemia Unawareness Using a Factory-Calibrated Continuous Glucose Monitoring System

Background

Approximately 90% to 95% of insulinomas are benign tumors, and the 5-year survival rate after resection is expected to be 95% to 100% [1]. Large series in the past have reported recurrence in 3.0% to 5.4% of cases [2]. Differentiation between benign and malignant tumors can be difficult and is based on a combination of preoperative and intraoperative findings [1]. Histopathological markers that are associated with the prognosis of insulinoma include the Ki-67 growth index, tumor size, vascular invasion, and local invasion [3]. The treatment of insulinoma is mainly surgical, but surgery should be performed only once a diagnosis has been confirmed; a blind pancreatectomy is not an appropriate treatment at this time. In addition, drugs such as somatostatin analogs and diazoxide that inhibit the biosynthesis and release of insulin can be used in cases in which a lesion cannot be identified after various imaging tests or when surgical treatment is difficult because of financial circumstances.

The continuous glucose monitoring (CGM) system was first certified by the US Food and Drug Administration in 1999 and has benefited many people with diabetes over the past 2 decades, as considerable technological advances have led to regulatory approvals for multiple continuous and semi-continuous blood glucose (BG) monitors [4]; CGM is also reportedly useful for determining the presence of hypoglycemia and for the application of drug therapy in patients with insulinoma [5]. The glucose is measured in the interstitial fluid glucose (ISFG) throughout the day using the glucose oxidase reaction via fluorescence or intravenously via mid-infrared, fluorescence, electrochemical, or electrochemical impedance spectroscopy. Then, it is converted to a BG concentration through a calibration process using a user’s frequent self-monitoring of BG levels [4]. The Freestyle Libre (Abbot Diabetes Care, Oxon, UK), a factory-calibrated CGM system, is the first glucose monitoring device that does not require calibration, with the added benefit of similar performance in both accuracy and BG control of CGM devices that require calibration at least twice a day [6]. CGM measures the glucose concentration in ISFG throughout the day and is widely used to manage BG levels in patients with diabetes mellitus; it is also reportedly useful for determining the presence of hypoglycemia and the application of drug therapy in patients with insulinoma. Except for some devices utilizing real-time CGM systems, however, most currently available CGM systems require frequent self-measurements and the calibration of capillary BG levels, limiting monitoring to the retrospective identification of past hypoglycemic events. Several devices have been developed to measure the glucose concentration in ISFG. For example, the Abbott FreeStyle Libre Pro (Abbot Diabetes Care), a factory-calibrated CGM system intended for professional use only, can monitor the ISFG for 14 consecutive days per sensor, longer than other devices.

Here, we report a patient with an insulinoma in the pancreatic uncinate process who was treated using diazoxide and an Abbott FreeStyle Libre Pro factory-calibrated CGM system until surgery.

Case Report

INVESTIGATION:

Laboratory test results revealed an HbA1c of 4.4% (28 mmol/mol), fasting BG level of 68 mg/dL, and a high fasting insulin level of 8.9 μU/mL, despite hypoglycemia. His insulin antagonist hormone levels (ACTH, 12.07 pg/mL; cortisol, 2.6 μg/dL; TSH 4.09 μIU/mL; FT4 1.23 ng/dL) were in the reference range, and the anti-insulin antibody test result was negative. The results of liver function tests were an AST/ALT of 38/55 IU/L and a γGTP of 84 U/L, and his kidney function test results were in the reference range. The endocrinological and imaging tests ruled out multiple endocrine neoplasia type 1 (intact PTH, 45 pg/mL; calcium, 9.2 mg/dL; phosphorus 2.9 mg/dL; GH, 1.74 ng/mL; IGF-1, 182 ng/mL), adrenal insufficiency, and hypothyroidism. None of the 3 indicators (ie, the Fajans index, Grunt index, or Turner index) for relative hyperinsulinemia were met, but the Endocrine Society index (BG <55 mg/dL; immunoreactive insulin [IRI] ≥3 μU/mL) was met, and a fasting test was performed. After 11 h of fasting, the test results continued to show a BG level of below 55 mg/dL, and at 19 h, the BG level was 50 mg/dL, so the fasting test was terminated. At that time, the plasma IRI level was 4 μU/mL and the C peptide level was 0.6 ng/mL. Since an insulinoma was suspected, an abdominal CT examination was performed and a suspected mass lesion (22×15 mm) was newly identified in the pancreatic uncinate process, with enhancement that was approximately equal to that of the pancreatic parenchyma. Although this lesion was not visible on the magnetic resonance imaging (MRI) findings that had been obtained 1 month earlier, a subsequent MRI of the abdomen showed a low signal during T1, a faint high signal during T2, and a high signal during diffusion-weighted MRI in the same region, indicating a relatively strong enhancement effect from the early phase (Figure 1). Selective intra-arterial calcium injections showed a significant increase in IRI only in the common duct of the inferior pancreaticoduodenal artery and the first jejunal artery. Hypoglycemia was observed throughout the day using the factory-calibrated CGM system, with a 24-h BG range of 45 to 61 mg/dL. No hypoglycemic symptoms were observed after selective intra-arterial calcium injections, even when the BG levels were between 45 and 61 mg/dL.

TREATMENT:

The decision was made to treat the patient medically until surgery. After starting diazoxide at a dosage of 75 mg/day (1.05 mg/kg); the hypoglycemia improved (Figure 2), and no adverse effects were observed. The patient subsequently underwent a laparoscopic pancreatic tumor (19×17×16 mm) enucleation.

A pathological examination revealed a well-differentiated neuroendocrine grade 2 tumor with a Ki-67 percentage of 3% (Figure 3). The hypoglycemic periods resolved after the surgery without the continued use of diazoxide.

OUTCOME AND FOLLOW-UP:

At the 1-month postoperative follow-up, the range of BG levels was 80 to 182 mg/dL, and there were no episodes of hypoglycemia after an overnight fast. We monitored his glucose levels every month for 6 months, and then every 6 months thereafter as an outpatient, and he had not developed any hypoglycemia, even after exercising, and his weight had not changed significantly.

Discussion

Several reports have described the use of CGM during the diagnosis and resection of insulinoma [7]. In the present case, we successfully avoided hypoglycemia in a patient with insulinoma with postprandial hypoglycemia by monitoring his ISFG levels using the Abbott FreeStyle Libre Pro, a factory-calibrated CGM system, before and after treatment with diazoxide. The factory-calibrated CGM was useful for determining the treatment efficacy of diazoxide therapy and for dose adjustments, and its future use for such purposes is expected. Since delays in responding to central nervous system symptoms caused by hypoglycemia can lead to brain damage and even death, managing patients in a timely and safe manner is crucial once an insulinoma is suspected; in reality, however, delays in the detection and diagnosis of insulinomas are common. In terms of detection, more than 80% of patients with insulinoma present with fasting hypoglycemia, which makes a diagnosis difficult in patients who present with only postprandial hypoglycemia [8], as in our case. Furthermore, repeated episodes of hypoglycemia can lead to “hypoglycemia unawareness”, which prevents early measures from being taken to prevent any worsening of the hypoglycemia and makes it harder to detect.

Confirming the presence of hypoglycemia arising from an insulinoma is often difficult [9]. In terms of the time required to obtain a diagnosis, as many as 50% of patients are diagnosed 5 years after the onset of symptoms [10], and about 20% are misdiagnosed as neurological or psychiatric disorders until the time of insulinoma diagnosis [8]. The classical diagnosis of insulinoma is based on meeting the criteria of Whipple’s triad: central nervous system symptoms during fasting, a plasma glucose (PG) level <50 mg/dL at the onset of symptoms, and the relief of symptoms with glucose administration. There are also 3 assessments of insulin hypersecretion during fasting: the Fajans index (IRI/PG), the Grunt index (PG/ IRI), and the Turner index (IRI×100/[PG-30]). However, none of these 3 indices, or other imaging modalities, such as angiography and dynamic CT, are sufficiently sensitive for the diagnosis of insulinoma [9]. Image processing techniques are often unable to detect a significant percentage of insulinomas (CT, 55%; MRI, 42%) [8] because of the characteristics of insulinoma: 81.1% of cases are less than 2 cm in size, the metastasis rate is 6.6%, and 12.6% of cases have multiple lesions [11]. In adults with symptoms of neuroglycopenia and confirmed hypoglycemia, the criterion standard for biochemical diagnosis is the measurement of PG, insulin, C peptide, and proinsulin levels during a 72-h fasting period; according to Service et al [12], up to 99% of insulinomas can be detected using this prolonged fasting test. Of note, however, the absolute insulin level does not necessarily increase in all patients with insulinoma; therefore, even normal insulin levels do not rule out the disease [3].

Over the past decade, CGM manufacturers have devoted much effort to overcoming the problem of inadequate accuracy and inconvenience for users [6]. Although the reading by the Abbott FreeStyle Libre Pro we used is highly accurate without calibration, and the operation is user-friendly, it does not have an alarming function for the out-of-range glucose values [13]. On the other hand, other recent monitoring systems, including the Abbott Freestyle Libre and Dexcom G6, are equipped with such a function and are capable of wireless communication [6].

Conclusions

In summary, we report a case of insulinoma with hypoglycemia unawareness in which appropriate treatment was facilitated using a factory-calibrated CGM system. This system is useful in patients with insulinoma for the detection of hypoglycemia and hypoglycemia unawareness and for monitoring the response to medical treatment, as well as for confirming a complete cure after surgery.