A follow-up study of fulminant type 1 diabetes mellitus

Objectives explore the possible mechanisms of glycol-metabolism and islet function in patients with fulminant 1 diabetes (FT1DM). with the progression of the disease, metabolic disorders and stress responses were relieved at the later stage. Viral infections (herpes simplex virus, Coxsackie virus), HLA-DQ, DR genes, GAD-Ab and other related antibodies may be involved in the occurrence of FT1DM.


Introduction
According to the etiological classi cation criteria of diabetes established by WHO in 1999, type 1 diabetes mellitus (T1DM) can be divided into two types: autoimmunity (type 1A) and idiopathy (type 1B) [1].
Fulminant type 1 diabetes mellitus (FT1DM) is a special subtype of T1DM. FT1DM is extremely rare, prevalence of FT1DM in Japan was revealed as 1.0-8.9% of T1DM and 0.1-0.2% of all types of diabetes [1,2]. It was rst proposed by Imagawa [1] in 2000. It has the characteristics of acute onset, dangerous condition, ketoacidosis (DK) or ketoacidosis (DKA), high blood glucose, low glycated hemoglobin (HbA1c), and severe islet failure [1]. It is temporarily classi ed as type 1B diabetes. In the recent years, FT1DM has attracted worldwide attention with the increasing number of cases. There are more reports of FT1DM, especially in East Asia [2][3][4][5][6]. These reports mainly described the clinical characteristics and etiological analysis of FT1DM. However, there are few follow-up studies on FT1DM. At present, the pathogenesis of FT1DM has not been fully understood. Most studies suggested that viral infection and human leukocyte antigen (HLA) gene might be related to its pathogenesis [7]. However, there were few reports about the viral infection in FT1DM patients and the correlation of HLA-DRB1, DQA1 and DQB1 genes in Chinese population.
We studied 13 FT1DM in-patients from September 2000 to March 2015to investigate the changes of glucose metabolism and islet function in patients with fulminant type 1 diabetes mellitus (FT1DM) over time and possible underlying mechanisms.

Subjects
The protocol and informed consent document were approved by the Institutional Ethics Committee at Nanjing First Hospital. All patients subsequently gave written informed consent. 1. Inclusion criteria: Diagnosis of FT1DM was based on the diagnostic criteria proposed by Imagawa et al.
[8] : 1) ketosis or ketoacidosis within a week of hyperglycemia; 2) initial blood glucose > 16 mmol/L (> 288 mg/dl) and glycosylated hemoglobin < 8.7%; The abdominal plasma C-peptide was < 0.1 nmol/L (< 0.3 ug/L) at onset and < 0.17 nmmol/L (< 0.5 ug/L) after stimulation (after meal or glucagon injection). 2. Exclusion criteria: Complicated with severe acute myocardial infarction, acute pancreatitis, primary liver disease and kidney disease; other mental illness that the researchers considered unsuitable for participation. 3. Thirteen FT1DM patients with ketoacidosis as the rst symptom were enrolled into Department of Endocrinology, Nanjing First Hospital from September 2000 to March 2015. All FT1DM patients were followed up from March 2015 to March 2018.

Methods
Data collection: All the participated subjects were lled out by the residents of Endocrinology Department in a uniform form of questionnaire, which was checked by the quality supervisor and entered uniformly.
For further diagnosis of molecular genetics, 2ml of peripheral blood samples for the patients were collected. The genotypes of HLA-DRB1, DQA1 and DQB1 were ampli ed and analyzed by polymerase chain reaction (PCR) ampli cation with sequence-speci c primers (PCR-SSP) as previous reported 54 pairs primers [9,10].The primers were synthesized by Shanghai Genomics institution (Shanghai, China). PCR Products were identi ed by DNA Agarose Gel Electrophoresis.
Analyses were performed using the SPSS 22.0 (SPSS, Science, Chicago, USA) statistical package. All variables were tested for normal distribution of the data. The normal distribution data were analyzed by the t-test and presented as means ± SD. The data of non-normal distribution were tested by Mann-Whitney test and expressed as median (IQR). Chi-square test was used to compare the rates between two groups. All comparisons were 2-sided at the 5% signi cance level. P value < 0.05 was considered to be statistically signi cant.

The follow-up results of 13 FT1DM
The average age of onset was (46.54 + 7.14) years. The average course of disease was 7.38 ± 4.89 years in general data. Compared with the baseline (at the time when patient was diagnosed as F1TM), the waist-hip ratio signi cantly increased during follow-up period (P < 0.05). There was no signi cant difference among the pre-meal insulin dose, basal insulin dose and daily total insulin dose (u/d/Kg) (P > 0.05). (Table 1)
In this study, 13 patients with FT1DM were followed up for an average of 7.38 years. Patient WBC, AST, K, Cr, GGT, CK and AKP follow-up were signi cantly lower than those at onset of FT1DM, while TC and HDL were signi cantly higher than those in FT1DM follow-up, and tended to normalize, indicating that acute metabolic disorders recovered over time.
The results showed that the blood glucose of FT1DM patients was poorer, the function of islets was worse, and HbA1c was signi cantly higher compared than those at the onset stage of the disease. FT1DM had the characteristics of higher blood glucose and lower HbA1c at the onset stage of the disease. With the prolongation of the course of disease, the increased HbA1c indicated that the control of blood glucose was still unsatisfactory even after insulin treatment in the later stage of the disease. C-P and C-P 120 were signi cantly lower than those at the onset of FT1DM, indicating that islet beta cells did not completely damaged and they progressively reduced function after acute phase. It is different from our previous knowledge that the islet function of FT1DM patients was almost lost at onset of the disease [1,4,6]. This may be related to the short follow-up time and limited case reported.
A short-term follow-up of FT1DM patients by Imagawa [11] et al., showed that there was no difference in HbA1c between FT1DM patients and T1DM patients in 3, 6, and 12 months of follow-up. However, a 5year follow-up study of FT1DM patients by Murase et al. [12] indicated that C-P120 maintained a low level for 5 years and C-P120 was lower in FT1DM patients than in T1DM patients. Furthermore, Tang et al., reported ve FT1DM patients with follow-ups of 52 months [13]. The study showed that the HbA1c level of the F1DM patients increased and the islet function of all patients was worse than at the onset of the disease even after intensive treatment. This is similar to the current study.
Lan Liu et al. showed that HbA1c was elevated after 6 months of follow-up of FT1DM patients and C-P0 and C-P120 were not signi cantly different from onset of the disease [5]. 5 FT1DM patients were followed up by Lu Zeyuan et al. for 3-26 months indicated that C-P0 and C-P 120 were extremely poor, and there was no difference with the onset of FT1DM [14]. 6 patients were followed up with FT1DM for 9-72 months by Fan Yujuan et al. [15].They found that C-P0 and C-P120 were close to the level at admission, suggesting that pancreatic beta cells were completely and irreversibly destroyed. Huang Huibin et al., reported 2 patients and followed up the function of their pancreatic beta cell at 1 and 7 months respectively [16]. The results showed no improvement in the function of their pancreatic beta cell and the second case showed the function of islet B cells decreased signi cantly in the later stage. But our study showed followed-up C-P0 and C-P120 were signi cantly decreased compared with baseline.
At present, the pathogenesis of FT1DM is not completely clear. The pathogenesis of FT1DM may be mediated by many factors, including viral infection, pregnancy, drugs, autoimmune and genetic factors [17]. A study in Japan reported that the onset of the disease was related to genetic background and viral Studies suggested that enteroviruses and chemokines not only destroyed the islet beta, but also further accelerated the autoimmune response mediated by remaining islet beta cells until all function of islet beta cells was destroyed [20,21]. Our study shows that there is a high positive rate of viral antibody titer detected in the non-acute stage of FT1DM, which may support the above pathogenesis mechnism.
Imagawa et al. showed that GAD-Ab and INS-Ab were detected in FT1DM patients with low titer and short duration [11], while the ICA-Ab did not appear in FT1DM [4]. In the current study, it was found that GAD-Ab in the FT1DM patients was initially 20% and decreased to 7.7% during follow-up, while INA-Ab increased from 30.8% initially in the early stage of FT1DM to 38.5% during follow-up, and ICA-Ab was detected in 11.1% FT1DM patients during follow-up. These ndings are quite surprising contrast with Imagawa's study, which showed only 4.8% (7/138) patients had GAD-Ab detected in a national survey in 2003 [12]. However, they are consistent with Zheng or Liu's studies that detected GAD-Ab in 35% and 30%, respectively, of FT1DM patients [2,5].
The increased INA-Ab was probably due to the use of insulin. These results showed that the persistence of multiple antibodies in FT1DM from the initial stage of onset to several years later might be one of the reasons for the progressive decline and irreversible function of beta cells. In this study, multiple antibodies detected may support the autoimmune involvement in the occurrence of FT1DM [2].
In summary, with the progress of the disease, the function of FT1DM islet beta cells showed a trend of progressively irreversible destruction; various metabolic disorders and stress reactions of FT1DM were alleviated at the later stage of onset. Viral infection (herpes simplex virus, coxsackievirus), virus antibody production, HLA-DQ, DR gene, GAD-Ab and other related antibodies may be involved in the occurrence of FT1DM.