ACRT Journal

Introduction

Hypoglycemia is a rare condition in individuals not using medications that alter glucose levels. Therefore, its diagnosis should be confirmed using Whipple’s triad (low glucose levels, accompanied by compatible symptoms that improve with glucose correction), followed by further investigation with the fasting test [1], which allows for the classification of hypoglycemia as either hyperinsulinemic or non-hyperinsulinemic.

One cause of hyperinsulinemic hypoglycemia is insulinoma, a rare, insulin-secreting pancreatic neuroendocrine tumor that is most often benign, solitary, sporadic, well-differentiated, and exhibits a slight female predominance [2,3]. However, careful evaluation of additional fasting test parameters, beyond insulin levels, is essential to exclude, among other causes, factitious hypoglycemia.

Factitious hypoglycemia refers to the intentional induction of low blood glucose levels to assume the role of a sick individual, done consciously and deliberately, and may occur even without any apparent external gain, with the primary aim of seeking medical attention. Munchausen syndrome is a severe, chronic form of factitious disorder, characterized by repeated hospitalizations and an extensive medical history, including unnecessary diagnostic and therapeutic procedures [4]. This disorder differs from malingering, where the motivation is financial gain or the avoidance of legal responsibilities, and individuals typically seek to avoid invasive procedures or therapies [5].

We present a case of factitious hypoglycemia that occurred following successful surgical treatment of an insulinoma, emphasizing key aspects of the diagnostic evaluation to minimize or prevent harm from unnecessary invasive medical interventions. To the best of our knowledge, no similar case has been reported to date.

Case Presentation

A female patient with a past history of polycystic ovary syndrome (PCOS), without the use of medications, experienced recurrent episodes of adrenergic and neuroglycopenic symptoms starting at the age of 14, occurring primarily during fasting, particularly in the morning, with improvement after eating. These episodes were not investigated at that time.

At the age of 27, she was admitted to the emergency room following a seizure due to a hypoglycemic episode (capillary blood glucose < 55 mg/dL). Upon admission, Whipple’s triad was confirmed.

Seen the long history of hypoglycemic episodes, the hypothesis of insulinoma was raised and an investigation was carried out with the fasting test, which revealed hyperinsulinemic hypoglycemia (glucose 53 mg/dL, normal value > 55 mg/dL; insulin 11.5 µU/ mL, normal value < 3 µU/mL) and elevated C-peptide (3.83 ng/ mL, normal value < 0.6 ng/mL) (Table 1).

The patient had normal calcium and pituitary hormone levels, with no dermal lesions such as trunk collagenoma or nasal angiofibroma. Her personal and family history was negative for tumors associated with genetic syndromes, including multiple endocrine neoplasia type 1 and tuberous sclerosis.

During hospitalization, the patient developed Raynaud’s phenomenon, leading to the diagnosis of Raynaud’s disease, with no additional symptoms and signs suggestive of autoimmune disease.

Imaging studies were conducted to investigate a pancreatic tumor, and abdominal magnetic resonance imaging (MRI) revealed a 1.4 cm nodule in the pancreatic tail (Figure 1). Laparoscopic distal pancreatectomy was performed, confirming the diagnosis of insulinoma via immunohistochemistry (IHC) (Figures 1, 2, 3).

She was discharged four days after surgery, as there was complete remission of hypoglycemia during the perioperative period, without complications. Hypoglycemic episodes recurred two weeks later, and the patient was admitted for further investigation.

Imaging studies did not reveal any new pancreatic lesions, although 68Gallium-DOTATOC PET/CT suggested a lesion in the head of the pancreas, located near a region of physiological capture, without anatomic correspondence by aggregated computed tomography (Figure 4).

Somatostatin analog therapy was initiated, but there was no significant improvement after several weeks, leading to discontinuation two months later. Due to recurrent hypoglycemia and new seizures, the patient required continuous intravenous glucose infusion through central venous access and prolonged hospitalization.

Due to Raynaud’s disease, autoimmune hypoglycemia was excluded based on negative anti-insulin and antinuclear antibody tests. New insulin and C-peptide samples were collected during hypoglycemic episodes, but the results were inconclusive, with a borderline C-peptide value during hyperinsulinemic hypoglycemia, both with and without somatostatin analog therapy (Table 1). Factitious hypoglycemia was then suspected, and the patient was kept in an isolated bed for observation, continuing to experience hypoglycemic episodes.

Due to the inconclusive diagnosis, the patient was referred to our tertiary center for further investigation of insulinomatosis, nesidioblastosis, occult malignant insulinoma, or multiple insulinomas associated with inherited disorders. A new fasting test was again inconclusive (Table 1). Other attempts to localize a potential occult tumor using MRI and endoscopic ultrasound were unsuccessful. While further fasting test, measurements were being conducted and the pathology of the surgical specimen was reviewed, the patient experienced recurrent hypoglycemic crises during hospitalization, requiring intravenous glucose infusion through central venous access. As a result, pancreatic arteriography was performed, followed by a selective arterial calcium stimulation test, which returned negative results.

Shortly thereafter, the complete results of the fasting test revealed hyperinsulinemic hypoglycemia (glucose 42 mg/dL, insulin 43.3 mcU/mL) with inappropriately normal values for C-peptide (0.5 ng/mL, normal value < 0.6 ng/mL) and proinsulin (2.39 pmol/L, normal value < 5 pmol/L), negative ketonemia, and a positive response to glucagon, suggesting malicious and surreptitious insulin administration (Table 1). In parallel, insulinomatosis and nesidioblastosis, which had also been considered potential etiologies, were definitively excluded after a thorough review of the surgical specimen (Figure 3).

The patient remained hospitalized for 10 months across various hospital services, displaying a strong inclination to assist the nursing team with organizing supplies, independently adjusting the glucose infusion rate, refusing medication, and engaging in conflicts with other patients. Both the psychology and psychiatry teams were involved in her care.

Due to suspicious behaviour, a review of her personal belongings revealed vials of human insulin and injection needles. The patient admitted to injecting insulin to simulate symptoms following pancreatectomy, also disclosing a history of childhood trauma, sexual abuse, and previous suicide attempts.

Following psychiatric evaluation, a diagnosis of borderline personality disorder was made, and Munchausen syndrome was diagnosed. She initiated treatment with fluoxetine and quetiapine. The patient was discharged after securing a support group and arranging psychiatric follow-up. At the one-year follow-up, she missed several scheduled appointments and was readmitted to the emergency department six months ago due to a new hypoglycemic episode. Since then, no further hypoglycemic events have occurred. In her most recent consultation, she reported that she was enrolled in a nursing technical course and was performing excellently. Genetic testing for MEN1 was performed, and no pathogenic variants in the MEN1 gene were identified by Sanger sequencing or Multiplex Ligation-dependent Probe Amplification (MLPA) assay.

Table 1: Typical biochemical pattern of endogenous hyperinsulinemic hypoglycemia documented by the fasting test performed prior to partial pancreatectomy for insulinoma, followed by an erratic and inconclusive biochemical profile due to early recurrent postoperative hypoglycemia, ultimately leading to laboratory findings strongly suggestive of factitious hypoglycemia, likely resulting from surreptitious exogenous insulin use.

Figure 1: Preoperative radiological image and macroscopic view of the pancreatic tumor resected during subtotal pancreatectomy. A. Abdominal magnetic resonance imaging: A nodular formation (yellow arrow) at the body-tail junction of the pancreas, measuring 1.4 x 0.9 cm; B. Surgical specimen following partial pancreatectomy, showing a solitary pancreatic nodule at the body-tail junction. The nodule, sectioned in half (blue arrows), is wine-colored with well-defined borders, and measures 1.5 x 1.3 x 1.0 cm.

Figure 2: Histology of the pancreatic neuroendocrine tumor (NET) located at the body-tail junction. A-E: Stained with hematoxylin and eosin. A. Microscopic panoramic view of the NET measuring 1.5 x 1.3 x 1.0 cm, unifocal, well-demarcated, with clear borders free of neoplasia, and no evidence of angiolymphatic or perineural invasion, nor lymph node involvement (0/7) (scale: 500 µm). B. Transition between tumor and normal pancreatic tissue, with the NET on the left and normal tissue on the right (scale: 200 µm). C. Normal pancreatic tissue, including multiple endocrine cell islets (red arrows) (scale: 100 µm). D. Tumoral tissue (scale: 100 µm). E. Tumoral tissue at higher magnification (scale: 50 µm).

Figure 3: Immunohistochemistry (IHC) demonstrating an insulin-secreting grade 1 (G1) well-differentiated neuroendocrine tumor (WDNET). A-C. Positive IHC for cytokeratin (CK8/18) (A), chromogranin (B), and synaptophysin (C), classifying the tumor as neuroendocrine (positive for AE1/AE3, not shown). D. Ki67 labelling index < 3%, classifying this WDNET as G1. E-F. Positive IHC for insulin in the neuroendocrine tumor (E) and in normal islet cells (F). G-H. Negative IHC for glucagon in the insulin-secreting NET (G) and positive in normal islet cells (H). I-J. Negative IHC for gastrin (I) and somatostatin (J). Scales in µm.

Figure 4: False-positive results suggesting a pancreatic tumoral lesion despite an inappropriately normal C-peptide value, which was borderline for the laboratory diagnosis of hyperinsulinemic hypoglycemia (see Table 1). A-C. 68Gallium-DOTATOC PET/CT showed a focal area of uptake (SUV max = 6.9) in the uncinate process/head of the pancreas (red circle); D. with no corresponding anatomical findings on computed tomography (blue circle) and another area of uptake near the site of previous surgical manipulation (SUV max = 4.9) (not shown). MRI suggested a possible 3-mm lesion in the same location within the uncinate process (not shown). Endoscopic ultrasound was normal in this region, and biopsy identified an inflammatory lymph node near the area of prior surgery (not shown). Given the inconclusive results of subsequent biochemical tests, pancreatic arteriography was performed with a selective arterial calcium stimulation test, which yielded a negative result (not shown).

Discussion

We reported a case diagnosed with failure of the resolution of hypoglycemia after distal pancreatectomy for insulinoma in a young adult woman. Surgery is the cornerstone of treatment of insulinoma and scenarios of unsuccessful surgical treatment occur exceptionally. In this context, the main causal events rememorated by specialists associated with primary surgical failure are represented for inherited syndromic forms of insulinoma, especially by multiple endocrine neoplasia type 1 and exceptionally by tuberous sclerosis, by insulinomatosis and by nesidioblastosis, malign insulinoma or by initial failure of radiological localization of solitary insulinoma. Ultimately, in a case with a previous background of proved organic, the hypothesis of factitious hypoglycemia may be underestimated. Thus, this case teaches us that the basic rules of investigation in cases with recurrent hypoglycemia must be fully remade step by step.

Occult insulinoma was the first differential diagnosis to be excluded in the present case as insulinoma was found with complete remission of hypoglycemia up to 2-3 weeks after surgery. Thus, inherited disorders frequently related with multifocal insulinoma were investigated and appropriately excluded. Around 5% to 10% of insulinomas are associated with multiple endocrine neoplasia type 1 (MEN1) syndrome [3,6,7]. The youth and recurrence of hypoglycemia were points favouring this hypothesis [6,7]. However, the absence of primary hyperparathyroidism observed in at least 80% of MEN1 cases at 30y, the absence of combination of other MEN1-related tumors, especially the frequent combination of non-functioning pancreatic neuroendocrine tumors with insulinoma, absence of MEN1related dermic lesions, negative familial history and, absence of germline MEN1 mutation excluded MEN1 [8,9].