IGF-2-mediated resistant hypoglycemia in a patient with a nonislet cell tumor

Dear Editor:

The etiological diagnosis of recurrent hypoglycaemia represents a real challenge for the clinician. There are a multitude of causes that can lead to hypoglycaemia: hypoglycaemic drugs, malnutrition, sepsis, liver disease, etc. Neoplasm-associated hypoglycaemia is a separate entity, which may also be caused by different mechanisms: production of insulin, IL-1, IL-6, destruction of the liver or adrenal glands, etc. In all cases, the patient must be assessed in a comprehensive way and any potential causes of hypoglycaemia must be considered, without overlooking those that are uncommon. The aim of this paper is precisely to take into consideration one of these low-frequency causes that should be considered in the differential diagnosis: overproduction of IGF-2, known as Doege–Potter syndrome. The estimated incidence is 1 case per million population, with very few cases reported worldwide. This paper describes the case of a patient treated in our department. It includes how to reach the diagnosis and the effective treatment available, according to the latest available evidence.

We report the case of an 83-year-old male patient with a history of chronic obstructive pulmonary disease and diagnosis of a solitary fibrous tumour of the pleura 3 years earlier. The patient refused surgical intervention for the tumor, so he continued follow-up by the Pneumology Department.

His family doctor detected fasting glucose levels <70 mg/dl in the control tests, which had not caused any symptoms. However, sometime later, the patient began to present with numerous episodes of hypoglycaemia together with neuroglycopenic symptoms that required emergency care. Given the frequency and poor response to hygienic-dietary measures, the patient was admitted for further evaluation. Hypoglycaemia, with fasting glucose levels below 50 mg/dl, was confirmed with blood tests. Coinciding with these, the patient showed symptoms of hypoglycaemia that disappeared with glucose intake, thus fulfilling Whipple's triad.

The laboratory tests on admission showed normal renal and hepatic function values. The patient's ingesta was normal. On admission, he received glucose solution and a diet rich in complex carbohydrates administered abundantly and frequently. Even with these measures, he continued to require frequent use of 50% glucose solution as rescue to correct episodes of hypoglycaemia. A blood test was performed in one of the hypoglycaemia episodes and the following results were obtained: C-peptide <0.2 ng/ml and insulin <3 µU/ml, with no oral antidiabetic agents being detected. These data, considering the patient's history, focused on hypoglycaemia due to non-islet cell tumour (NICT) as the most likely aetiology.

After the etiological work-up, empirical treatment with prednisone 20 mg/12 h was initiated and glucose solution was discontinued. With these measures, blood glucose remained in the normal range.

The following results were then obtained from a reference laboratory: IGF-1: 36 ng/ml (normal 55−166 ng/ml), IGF-2: 481 ng/ml (normal 350−481 ng/dl), IGF 2/IGF 1 ratio: 13.36 (normal < 10). These results confirmed that the origin of the recurrent hypoglycaemias was IGF-2 synthesis by the fibrous tumour of the pleura.

Given the characteristics of the patient and the size of the tumour, the Thoracic Tumour Committee decided to continue with conservative measures. Prednisone treatment was maintained, and the patient did not experience any further episodes of hypoglycaemia.

There are many NICTs capable of causing hypoglycaemia. They are generally large in size and are normally found in the thoracic or retroperitoneal region. If the hypoglycaemia is caused by a fibrous tumour found in the chest, it is called Doege–Potter syndrome. The solitary fibrous pleural tumour is the most common hypoglycaemia-producing NICT. The most common cause of hypoglycaemia due to NICT is tumour overproduction of an insulin-like factor, usually incompletely processed IGF-2, called pro-IGF-2, or more rarely, IGF-1.^2,3 IGF-2-producing NICTs develop this molecule with a higher molecular weight than usual. Although many cases of NICT have elevated IGF2 levels, cases with normal or low values have also been reported. This discrepancy is thought to be due to variability in laboratory measurement methods and their ability to detect abnormal forms of IGF-2. IGFs induce an increase in glucose utilisation and an inhibition of glucose release from the liver and ketogenesis. Pro-IGF-2 also suppresses the counter-regulatory hormones glucagon and GH. Approximately half of the cases of NICT present with hypoglycaemia as the first manifestation of the tumour. The other half have the tumour identified prior to the onset of hypoglycaemia.⁴

The diagnosis of NICT hypoglycaemia is based on clinical and biochemical findings. When a patient with a known tumour presents with hypoglycaemia, the cause is usually evident from the clinical history and examination. When the cause is not obvious, laboratory tests are needed. Initial laboratory evaluation includes measurement of glucose, insulin, proinsulin, C-peptide, beta-hydroxybutyrate, and sulfonylureas/meglitinides during a hypoglycaemic episode.^1,3 A low hydroxybutyrate level points to insulin or IGF hypoglycaemia. Insulin, C-peptide, and proinsulin levels are elevated in subjects with insulinomas, oral hypoglycaemic agents, and insulin autoimmunity. Sulfonylureas or meglitinides are present in hypoglycaemia mediated by these oral hypoglycaemic agents. In patients with exogenous insulin administration, plasma insulin levels are higher than in insulinoma, but C-peptide and proinsulin have low values. In contrast to individuals with hyperinsulinemic hypoglycaemia, patients with IGF hypoglycaemia have low levels of insulin and C-peptide during hypoglycaemia. When an IGF-producing NICT is suspected, GH, IGF-1, and IGF-2 are measured. GH levels are typically low (except in brief episodes of hypoglycaemia in which its production is stimulated). Depending on the IGF-2 measurement method, the levels may or may not be elevated, as discussed above. Although the levels are normal, IGF-1 will be suppressed below 100 ng/ml, so the IGF-2/IGF-1 ratio is high.

The finding of hypoglycaemia with low levels of insulin, C-peptide, and beta-hydroxybutyrate in an apparently healthy person should lead to a rapid tumour search by cross-sectional imaging of the chest, abdomen, and pelvis, because in most cases these tumours are located there. Dynamic tests can be performed in patients with suspected adrenal failure.

Logically, the main treatment to resolve hypoglycaemia is that of the tumour itself.^2,3,5 When a tumour is producing IGF, complete removal of the tumour cures hypoglycaemia. If the tumour is not resectable, a reduction of the tumour mass is usually performed. In addition, depending on the type of tumour, chemotherapy, radiotherapy, or selective embolization of the blood vessels that nourish the tumour may be used. If the tumour cannot be treated, glucocorticoids are initiated at equivalent doses of prednisone 30−60 mg/dl. These decrease the amount of macro IGF-2 by suppressing its production or increasing its clearance. In addition, they stimulate gluconeogenesis and decrease IGF-2 outflow from the endothelial barrier. If hypoglycaemia persists, continuous intravenous infusion of glucagon (0.06−0.3 mg/h) can be used; recombinant GH could also be used at supraphysiological doses of 3−12 mg/day. GH suppresses peripheral glucose uptake and decreases passage through the endothelial barrier of IGF-2, but its use may be limited by the need for high doses with side effects. Diazoxide and octreotide decrease the secretion of insulin by the pancreatic islet cells, so they are not useful in NICT, as hypoglycaemia is not due to hyperproduction of insulin by these cells.

In conclusion, in the care of patients with hypoglycaemia and neoplasms, a targeted history-taking and examination are particularly relevant, as well as a correct interpretation of laboratory findings, considering the possibility of less common aetiologies in the differential diagnosis, such as IGF-2 overproduction.

Conflict of interests

The authors declare that they have no conflicts of interest for this publication.