Changes in glucagon secretion induced by food intake in fulminant type 1 diabetes mellitus: a case report

Abstract

Fulminant type 1 diabetes mellitus (FT1DM) is a subtype of type 1 diabetes mellitus and is characterized by a remarkably abrupt onset and almost complete destruction of β-cells within a few days. Here, we report a case of diabetic ketoacidosis in a 63-year-old man with no history of hyperglycemia. The patient was diagnosed with FT1DM and had almost no insulin secretion. We examined his insulin and glucagon secretions induced by a liquid meal test at the onset of FT1DM and 1 year later. The results suggested severely attenuated insulin secretion and an undetectable level of serum insulin 1 year after onset. In contrast, glucagon secretion, which was highly impaired at onset, increased in response to food intake. Although previous reports have suggested that both β- and α-cells of pancreatic islets are damaged in patients with FT1DM, the number of α-cells may increase over time after the onset of FT1DM.

Introduction

Type 1 diabetes mellitus (T1DM) is caused by a decline in insulin secretion due to the destruction of pancreatic β-cells. Fulminant T1DM (FT1DM) is a subtype of T1DM, which is characterized by rapid and almost complete destruction of β-cells within a few days that leads to severe hyperglycemia and diabetic ketoacidosis (DKA), and patients generally test negative for islet cell-related autoantibodies. In Japan, FT1DM accounts for approximately 20% of cases of acute-onset adult T1DM with ketosis at disease onset [1]. Previous reports have suggested that both β- and α-cells of pancreatic islets are reduced in patients with FT1DM, suggesting impairments of both insulin and glucagon secretions [2, 3]. Plasma glucagon levels were evaluated and compared using a sandwich enzyme-linked immunosorbent assay (ELISA) at the onset of FT1DM and 1 year later.

Case report

A 63-year-old man was admitted to our hospital with a diagnosis of DKA. He had no relevant medical history or abnormality in a medical examination conducted in March 2020. He had no history of excessive soft drink consumption and had developed sudden fatigue, thirst, and polyuria 2 days before admission. His symptoms had progressively worsened, and he was referred to the hospital because of excessive fatigue and disturbed consciousness.

The patient had a blood pressure of 162/92 mmHg; pulse rate, 60 beats/minute; and body temperature, 36.6 °C. As shown in Table 1, he developed DKA, with a random sample plasma glucose level of 484 mg/dL, arterial blood pH of 7.225, bicarbonate level of 12.8 mmol/L, and 3+ urinary ketone bodies. Despite the presence of hyperglycemia, his glycated hemoglobin (HbA1c) level was within the normal range (5.7%), and the urinary C-peptide reactivity (CPR) was extremely low (1.0 μg/day). His fasting serum CPR was <0.2 ng/mL (reference range 0.8–3.2 ng/mL). Findings for anti-glutamic acid decarboxylase antibodies (reference range <5.0 U/mL) and anti-insulinoma-associated antigen-2 antibodies (reference range 0.0–0.5 U/mL) were negative. In addition, serum amylase, elastase-1, and lipase levels had increased to 176 U/L (reference range 44–132 U/L), 688 ng/dL (reference range <300 ng/dL), and 54 U/L (reference range 5–35 U/L), respectively. The patient was diagnosed with FT1DM. We administered intravenous saline and regular insulin to treat DKA at the time of admission, followed by initiation of subcutaneous insulin injection therapy with insulin lispro and insulin glargine U-300. The patient’s serum CPR was undetectable (<0.2 ng/mL) before and 6 min after intravenous glucagon loading, indicating FT1DM. Human leukocyte antigen class II genotypes were DRB1*09:01:02/15:01:01 and DQB1*03:01:01/03:03:02.

Table 1.

Laboratory data on admission and during hospitalization

CBC		CK	193 U/L	Diabetes
WBC	8300 /μL	AMY	176 U/L	PG	484 mg/dL
RBC	575×104 /μL	P-AMY	44 IU/L	IRI	< 0.4 μU/mL
Hb	17.7 g/dL	Elastase1	688 ng/dL	S-CPR	< 0.2 ng/mL
Ht	55.1 %	Lipase	54 U/L	HbA1c	5.7 %
Plt	22.1×104 /μL	Na	132.1 mEq/L	GA	19.5 %
Blood chemistry		K	5.79 mEq/L	GAD Ab	< 5.0 U/mL
AST	25 U/L	Cl	96.4 mEq/L	IA-2 Ab	< 0.6 U/mL
ALT	42 U/L	Ca	10 mg/dL	HLA genotype
LDH	214 U/L	P	6.5 mg/dL	DRB1*09:01:02/15:01:01,
γGTP	42 U/L	Coagulation		DQB1*03:01:01/03:03:02.
T-Bil	3.06 mg/dL	PT	12.3 sec	Arterial blood gas
ALP	312 IU/L	APTT	25.6 sec	pH	7.225
ChE	387 U/L	Urinalysis		pO2	75.4 mmHg
CRP	0.73 mg/dL	pH	5.0	pCO2	31.6 mmHg
BUN	32.4 mg/dL	Protein	1+	HCO3-	12.8 mmol/L
Cre	1.15 mg/dL	Glucose	4+	BE	− 14.8 mmol/L
UA	6.7 mg/dL	Ketone	3+	AG	22.9 mmol/L
TP	9.1 g/dL	Blood	±	Lactate	15.4 mg/dL
Alb	5.7 g/dL	U-CPR	1.0 μg/day

PG plasma glucose, IRI immunoreactive insulin, CPR C-peptide reactivity, HbA1c glycated hemoglobin, GA glycoalbumin, GAD Ab Anti-glutamic acid decarboxylase antibodies, IA-2 Ab Anti-insulinoma-associated antigen-2 antibodies, U-CPR urine CPR, S-CPR serum CPR, BE base excess, AG anion gap

On admission day 11, glucagon secretion induced by food intake was evaluated after the patient achieved a fasting plasma glucose level of <130 mg/dL. We assessed the secretion of glucagon at disease onset. For the liquid meal test, Peptamen AF^® (300 kcal/200 mL, 25% of the energy by proteins, 39% by fats, and 36% by carbohydrates; Nestle, Japan) was administered after the patient had fasted overnight. Peptamen AF^® was selected to induce glucagon secretion for the following reasons. First, the 75-g oral or intravenous glucose tolerance test elevates plasma glucose levels, which may influence glucagon secretion. Second, it contains more protein and less carbohydrates compared to other enteral nutrients at our hospital. We used it to prevent a rapid increase in blood glucose levels and examine glucagon secretion. Finally, hospital meals do not always contain consistent carbohydrates, proteins, or fats. Blood was collected before and 30, 60, 120, 180, and 240 min after the administration of Peptamen AF^® to measure plasma glucose, serum CPR, serum insulin, and plasma glucagon levels. Exogenous insulin injections were omitted in the morning and during the liquid meal test. We obtained written informed consent at the time of the clinical investigation, and the study was approved by the Hiroshima City Asa Citizens Hospital Ethics Committee.

Laboratory analyses were performed at SRL, Inc. (Tokyo, Japan). Plasma glucagon levels were measured using a specific double-antibody sandwich ELISA (Mercodia AB, Uppsala, Sweden/Cosmic Corporation Co. Ltd., Tokyo, Japan), according to the manufacturer’s instructions. Serum insulin levels were measured using the Lumipulse^® Presto Insulin kit (FUJIREBIO Co. Ltd., Tokyo, Japan). Serum CPR and plasma glucagon levels, which were in the undetectable range (<0.2 ng/mL and <3.5 pg/mL, respectively), were assigned values of 0.2 ng/mL and 3.5 pg/mL, respectively. Figure 1 shows the results, indicating that both serum insulin and plasma glucagon levels had considerably attenuated at the onset of FT1DM.

Fig. 1.

Line charts indicating plasma glucose level, serum C-peptide reactivity, serum insulin level, and plasma glucagon secretion in response to a liquid meal test in a patient with fulminant type 1 diabetes mellitus are shown. The dashed line shows the level of each factor at onset, and solid line shows the same 1 year later. Serum C-peptide reactivity and plasma glucagon levels in the undetectable range (< 0.2 ng/mL and < 3.5 pg/mL, respectively) have been assigned values of 0.2 ng/mL and 3.5 pg/mL, respectively. CPR, C-peptide reactivity

After discharge, the patient visited the hospital once a month, and his HbA1c levels fluctuated between 7.0 and 7.7%. We have started to use flash glucose monitoring for glycemic control [4]. Time in range(TIR, 70 to 180 mg/dL) fluctuated between 66 and 77%, time below range (TBR 54–69 mg/dL) was less than 2% and no severe hypoglycemic event. One year later, he was readmitted to our hospital to improve exacerbated postprandial hyperglycemia. In the repeat liquid meal test, although glucose elevation was similar to that at diabetes onset, the serum CPR was completely absent, consistent with the findings of the previous test. In contrast, glucagon secretion was markedly enhanced from 30 to 120 min after the administration of Peptamen AF^®. These results suggest that glucagon secretion in response to food intake increased, while insulin secretion remained absent 1 year after the onset of FT1DM.

Discussion

Here, we reported the change in glucagon secretion induced by food intake in a patient with FT1DM; it was highly impaired at onset and markedly enhanced 1 year later. FT1DM is a subset of idiopathic T1DM, characterized by remarkably acute and almost complete β-cell destruction and nearly no insulin secretion, even immediately after disease onset [5]. Previous reports have indicated that both β- and α-cells are reduced in number, suggesting that both insulin and glucagon secretions are impaired in patients with FT1DM [2, 3, 6]. Sayama et al. reported that β- and α-cell areas in islets were substantially decreased in patients with FT1DM compared to that in patients with autoimmune T1DM and control subjects [2]. Pancreatic biopsy sections from five patients with FT1DM, five patients with recent-onset typical autoimmune T1DM, and six controls were prepared and stained using the indirect immunoperoxidase technique to visualize pancreatic α- and β-cells. In patients with FT1DM, the area of β- and α-cells was considerably decreased compared to that in patients with autoimmune T1DM and controls. In contrast, the α-cell area was not considerably decreased in patients with autoimmune T1DM compared to that in controls. The immunohistochemistry analysis showed a lack of Fas and Fas ligand expression in FT1DM, suggesting that the mechanism of β-cell destruction in FT1DM differs from that in autoimmune T1DM. Komada et al. reported that glucagon secretion is impaired in patients with FT1DM [6]. They showed that arginine-induced glucagon secretion in patients with FT1DM, which was assayed using a radioimmunoassay kit, was attenuated compared to that in patients with T1DM, type 2 diabetes mellitus, or slowly progressive insulin-dependent diabetes mellitus. In our case, glucagon secretion in response to food intake was considerably attenuated at onset, in accordance with a previous report.

The use of streptozotocin to deplete insulin secretion in glucagon receptor knockout mice did not cause abnormalities in glucose tolerance [7]. In our FT1DM case, glucose levels were elevated during the liquid meal test, although both insulin and glucagon secretions were impaired. Hayashi et al. reported that an increase in GLP-1 production is responsible for sustained low blood glucose levels in the absence of glucagon signaling [8]. We could not evaluate the GLP-1 concentration in our FT1DM case. However, the patient’s blood glucose level was elevated during the liquid meal test, suggesting that GLP-1 secretion was considerably attenuated at the onset of FT1DM.

Previous reports have suggested that glucagon secretion may change over time. Kawamori et al. measured plasma glucagon levels in patients with T1DM and concluded that glucagon secretion was dysregulated [9]. In their study, glucagon secretion significantly correlated with the serum blood urea nitrogen (BUN) concentration but not with urine CPR levels or renal function, which suggests a strong link between glucagon secretion and amino acid metabolism. They also assessed the yearly change in plasma glucagon levels in the same patients [10]. The statistical correlation between glucagon levels at annual checkups and sustained significant correlation between glucagon and blood urea nitrogen concentrations suggest a constant dysregulation of glucagon in association with altered amino acid metabolism. They found that the plasma glucagon levels of patients with T1DM were positively correlated with the plasma glucose levels, contrary to the previous report, where the plasma glucagon levels were only significantly correlated with serum BUN levels. In our case, his BUN level was almost same 1 year after the onset. The changes in glucagon secretion might be involved in an instability of glycemic control. Insulin signaling in α-cells is an intrinsic cellular mechanism that regulates glucagon secretion. They concluded that in patients with T1DM, insufficient insulin action on α-cells, particularly the deficient intra-islet insulin from neighboring β-cells, may induce dysregulated glucagon secretion. In mice, α-cell hyperplasia has been observed in the endogenous pancreatic islets with hepatocyte-specific deletion of the glucagon receptor, and α-cell hyperplasia has been observed in settings of partial or complete glucagon deficiency or resistance to glucagon action [11]. In patients with FT1DM, depletion of pancreatic β-cells induces impaired insulin signaling in α-cells, which leads to dysregulation of glucagon secretion. Therefore, impaired hepatic glucagon signaling may stimulate α-cell hyperplasia, which leads to enhanced glucagon secretion over time after the onset of FT1DM. Under conditions of severe β-cell demise below a threshold, α-cell plasticity is stimulated, leading to α-to-β-cell conversion [12, 13]. Previous reports have suggested that β-cell dedifferentiation and conversion into α-cells were observed in mice with insulin-resistant diabetes. However, a remarkable effect of diet on β-cell dedifferentiation has been suggested in obese mice, suggesting a potential mechanism by which lifestyle changes act as an effective intervention against diabetes progression [14, 15]. Takahashi et al. reported the changes in glucagon secretion using the arginine stimulation test in the patient with FT1DM at the onset and 6 months later [16]. Contrary to our FT1DM case, they found the peak plasma glucagon levels decreased 6 months later, compared with those at the onset. In our case, we performed liquid meal test to stimulate glucagon secretion, using Peptamen AF®, which contains multiple amino acid such as glutamine, aspartic acid and so on. Glucagon secretion may differ depending on the type of amino acid and whether it is infused or oral.　Although pancreatic α- and β-cells are apparently damaged at the onset of FT1DM, the number of α-cells may increase because of pancreatic cell conversion or lifestyle improvement in the long term.

In conclusion, glucagon secretion induced by food intake was enhanced 1 year after onset, whereas it was highly impaired at the onset of FT1DM. Long-term longitudinal studies with a large sample size are desirable to confirm pancreatic endocrinological functions in patients with FT1DM.

Author contributions

NH and TM analyzed and interpreted patient data. NH performed data collection and wrote the manuscript with support from TM and MY. All authors have commented on previous versions of the manuscript. All authors have read and approved the final manuscript.

Declarations

Conflict of interest

Author Himeno, Matsuda, and Yoneda declare that they have no conflicts of interest.

Human rights statement

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and its later versions.

Informed consent

Informed consent was obtained from the patient for this case report.