A case of fulminant type 1 diabetes masquerading acute pancreatitis.

Fulminant type 1 diabetes is a novel and life-threatening subtype of type 1 diabetes characterized by its markedly abrupt onset and hyperglycemia with a relatively low glycated hemoglobin level at the onset. We herein describe a typical case of fulminant type 1 diabetes who presented with abdominal pain and vomiting, and whose laboratory test revealed elevated pancreatic enzymes. He was initially diagnosed with acute pancreatitis, but occult and prolonged hyperglycemia eventually led to a diagnosis of fulminant type 1 diabetes.

Case Presentation

A 36‐year‐old Japanese man presented at a local hospital with left upper quadrant abdominal pain, vomiting, and thirst which had started the previous day. He had no significant past medical history. He was 173 cm tall and weighed 67 kg, his BMI was 22.4 kg/m2, and there was no unintended weight loss. Although an abdominal ultrasound study was unremarkable, elevated pancreatic enzymes (Table 1), left upper quadrant abdominal tenderness on physical examination, and his history of taking excessive amounts of alcohol (60‐80 g of pure alcohol a day) led to a diagnosis of acute alcoholic pancreatitis. The patient refused hospitalization, against the physician's recommendation, because of his family situation. After 4 days, his symptoms such as abdominal pain and vomiting worsened, and he was admitted to the same hospital. Further laboratory tests (Table 2) revealed severe metabolic acidosis, positive urine ketones and glucose by dipstick test, hyperglycemia, and a relatively low HbA1c level. Despite his pancreas appearing normal on CT scan, he was diagnosed with acute pancreatitis (APACHE II score was 13) and pancreatic diabetes. Fasting and intravenous fluids, antibiotics, gabexate mesilate, and subcutaneous insulin injection therapy were started. Although his symptoms gradually improved, at his request he was transferred to our hospital 4 days after the admission to the local hospital.

Table 1

Laboratory findings on first visit to the previous hospital

Complete blood count		Blood chemistry
White blood cell	6780/mm3	AST	22 IU/L
Neutrophils	58.3%	ALT	20 IU/L
Lymphocytes	30.7%	LDH	156 IU/L
Hemoglobin	17.9 g/dL	γ‐GTP	32 IU/L
Hematocrit	49.9%	ALP	276 IU/L
Platelets	18.8 × 104/mm3	T‐Bil	0.52 mg/dL
		CPK	68 IU/L
		Alb	4.4 g/dL
		TG	108 mg/dL
		Na	139 mmol/L
		K	3.7 mmol/L
		Cl	103 mmol/L
		Ca	9.15 mg/dL
		BUN	8.2 mg/dL
		Cre	0.72 mg/dL
		Amylase	309 IU/L
		Lipase	793 U/L (17‐57)
		Elastase‐1	1221 ng/dL (<300)
		Trypsin	3619 ng/mL (100‐550)
		CRP	2.91 mg/dL

Table 2

Laboratory findings on admission to the previous hospital

Complete blood count		Blood chemistry
White blood cell	17 910/mm3	AST	31 IU/L
Neutrophils	88.7%	ALT	45 IU/L
Lymphocytes	7.4%	LDH	276 IU/L
Hemoglobin	17.8 g/dL	γ‐GTP	43 IU/L
Hematocrit	50.4%	ALP	423 IU/L
Platelets	28.0 × 104/mm3	T‐Bil	0.26 mg/dL
		CPK	53 IU/L
Urinalysis		TP	8.6 g/dL
pH	5.5	Alb	5.2 g/dL
Protein	−	TG	86 mg/dL
Glucose	4+	Na	129 mmol/L
Ketone bodies	3+	K	5.89 mmol/L
Occult blood	∓	Cl	96 mmol/L
		Ca	10.24 mg/dL
Artery blood gas on room air		BUN	21 mg/dL
pH	7.119	Cre	1.0 mg/dL
PaO2	106.2 mm Hg	Amylase	111 IU/L
PaCO2	25.3 mm Hg	CRP	1.25 mg/dL
HCO3	7.9 mmol/L	Blood glucose	433 mg/dL
Base excess	−21.5 mmol/L
Na	133.1 mmol/L
K	5.96 mmol/L
Cl	97.3 mmol/L

On admission to our hospital, as all of his symptoms apart from the thirst had disappeared and the levels of pancreatic enzymes had been normalized, we initiated oral diet and started preparing him for discharge. However, his hyperglycemia persisted, and he required approximately 0.3 U/kg insulin per day. On the 6th day following admission to our hospital, we conducted a glucagon stimulation test. His fasting serum C‐peptide level was 0.04 ng/mL, and 6 minutes after loading glucagon (1 mg) intravenously, it remained at 0.04 ng/mL. This indicated an absolute insulin deficiency, and after taking his clinical course and relatively low HbA1c level into consideration, he was diagnosed with fulminant type 1 diabetes. He was discharged after instruction about diabetes, diet, and insulin self‐injection.

Discussion

Fulminant type 1 diabetes is a novel subtype of type 1 diabetes which has been reported from Japan since the end of the twentieth century.1 As the name suggests, this subtype is characterized by its markedly abrupt onset with ketosis or ketoacidosis, but unlike other forms of ketosis‐prone diabetes, it is also characterized by hyperglycemia with a relatively low glycated hemoglobin level at the onset.2 According to its characteristics,3 fulminant type 1 diabetes is thought to meet the ADA/WHO criteria for type 1B diabetes. Although the pathogenesis of fulminant type 1 diabetes has not yet been elucidated, recent studies indicate the causes as either viral infection or genetic background, such as human leukocyte antigens.4

As the majority of the cases have been reported from Asia, especially from Japan, a nationwide survey was conducted in Japan which revealed some unique manifestations.3 First, 98% of the patients with this subtype showed increased levels of serum exocrine pancreatic enzymes. Contrary to the laboratory changes, most cases do not reveal remarkable findings in abdominal CT scans or abdominal ultrasound studies.4 Histological analyses revealed mononuclear infiltration of both the exocrine and endocrine pancreas,5 and it indicates the presence of both “microscopic” pancreatitis and insulitis as pathogenesis. Second, 72.5% of the patients with this subtype had abdominal symptoms such as nausea, vomiting, or abdominal pain.3 With these symptoms and elevated pancreatic enzymes, typical fulminant type 1 diabetes artfully masquerades as acute pancreatitis. In the absence of chronic pancreatitis nor evidence of massive destruction of pancreas, it may be premature to conclude that the hyperglycemia is caused by pancreatic diabetes. Rather, clinicians need to bear in mind, in such cases, that it could be the initial presentation of fulminant type 1 diabetes.

Unfortunately, our patient did not undergo a blood gas test, plasma glucose test, nor urine examinations on the first visit to the previous hospital. Then, he was initially diagnosed with acute pancreatitis based on abdominal symptoms and elevated pancreatic enzymes. To make matters worse, he had had alcohol‐related problems. At that time, hypothesis of pancreatitis might have been strongly suggested by his alcoholic background and the symptom of thirst might have been discredited by “confirmation bias.”6

On his admission day, although further laboratory tests had revealed hyperglycemia, severe metabolic acidosis, and positive urine ketones, he was diagnosed with acute pancreatitis and pancreatic diabetes. That decision might have been influenced by the initial diagnosis, and “pancreatic diabetes” might have been added as the reason for hyperglycemia. When the patient was transferred to our hospital, furthermore, our cognition and decision were also anchored by the former diagnoses. We also had tried to explain such unnatural features away by thinking “severe acute pancreatitis or chronic alcoholism sometimes causes metabolic acidosis” or “prolonged abdominal symptoms may cause fasting ketosis.” At that time, we seemed to be biased by “anchoring and adjustment.”7

Conclusion

Typical fulminant type 1 diabetes masquerades as acute pancreatitis. Not only understanding the characteristics of this subtype, it is important to actively seek for signs and symptoms such as thirst, polyuria, flu‐like symptoms,3 hyperglycemia, or ketoacidosis which indicate fulminant type 1 diabetes rather than acute pancreatitis.

Conflict of Interest

The authors have stated explicitly that there are no conflicts of interest in connection with this article.