📘 Viral and bacterial infections in type 1 diabetes

A virus can trigger the destruction of pancreatic beta cells through five main mechanisms. The first is molecular mimicry, which occurs when viral proteins resemble certain regions (autoantigens) of beta cells in structure. The immune system cells (T lymphocytes) initially activate against the virus and subsequently attack the beta cells as well. For example, a specific protein (2C) from Coxsackievirus B4 closely resembles the GAD65 regions of the beta cell. The second mechanism, direct cell destruction, involves infection of beta cells, viral replication inside the cells and their destruction. The third mechanism, bystander damage, involves the creation of a harmful proinflammatory environment around beta cells, which activates autoreactive T lymphocytes in the area, without cross-reactivity with the virus. The "cytokine storm" associated with SARS-CoV-2 infection could also fall under this mechanism [1].

The fourth mechanism, persistent infection, involves viruses establishing chronic infections in beta cells and pancreatic duct cells, with a low number of viruses in each cell. The chronic presence of viruses in these cells causes stress (at the level of the endoplasmic reticulum) and the release of cell fragments, which are irritating to the immune system. Coxsackievirus B can persist in beta cells, pancreatic duct cells, intestine and thymus, these serving as reservoirs for reinfection at any time in the future. The fifth mechanism, non-discriminatory (superantigenic) activation, involves the binding of virus fragments by T lymphocytes, which activate against a broad spectrum of potential enemies, among which (by mistake) are the beta cells [2].

The age at which a child encounters a viral infection is crucial for the risk of developing type 1 diabetes, because the first years of life represent a critical window of immune maturation. Prolonged Enterovirus B infections (not short, isolated ones) in the first years of life are associated with the development of islet autoimmunity. Respiratory infections, with the exception of a respiratory virus called adenovirus C, in the first six months of life significantly increase the risk of later developing autoantibodies, highlighting the extreme vulnerability of infants [3].

First viral infections at the age of 6-12 months were associated with a lower risk of type 1 diabetes, suggesting that early viral exposure can also have a protective, immune training effect, consistent with the hygiene hypothesis. Exposure to viral infections in early childhood, when the immune system is maturing, may help explain the fact that the incidence of type 1 diabetes is currently increasing most rapidly in children under five years of age [4].

Among all viruses studied, enteroviruses (especially Coxsackievirus B) are most strongly associated with type 1 diabetes. The rubella virus (in congenital infection) was associated with diabetes development in 12-20% of patients with congenital rubella syndrome. Vaccination programmes virtually eliminate this risk. SARS-CoV-2 is the third virus with significant evidence, with a 42% higher risk of discovering type 1 diabetes after COVID-19 infection in children [2].

Other associated viruses, but with more limited evidence, include rotavirus. There is molecular mimicry (similarity) between a rotavirus protein and certain small fragments (the IA-2 and GAD65 autoantigens) of pancreatic beta cells. Mumps virus has been historically associated with type 1 diabetes. MMR vaccination completely eliminates this risk. Epstein-Barr virus does not show clear evidence of association with type 1 diabetes, and cytomegalovirus has mixed evidence, with studies both for and against. Parechoviruses, parvovirus B19 and influenza viruses are currently being evaluated, with initial results suggesting an association, but possibly a weak one.

Enteroviruses, especially Coxsackievirus B, are considered the main environmental factors implicated in the onset of type 1 diabetes. The DiViD study (Diabetes Virus Detection), the first study to collect pancreatic tissue from living patients with newly diagnosed type 1 diabetes, detected live enteroviruses in the pancreas of all six patients studied (versus 2/11 controls). The TEDDY study, the largest prospective birth cohort study, analysed viruses from faeces (the faecal virome) and found an association between prolonged Enterovirus B infections (not short or isolated ones) and the development of islet autoimmunity [3].

Antiviral treatment for six months in children with newly diagnosed type 1 diabetes halves the rate of C-peptide decline (11% in the treated group versus 24% in the placebo group) [5]. Other prospective birth cohort studies that have consolidated the evidence include DIPP (Finland), DAISY (Colorado, USA), MIDIA (Norway) and BABYDIET (Germany). Children who later develop islet autoimmunity have a deficient antiviral immune activation upon enterovirus infection, suggesting that genetic susceptibility to type 1 diabetes also includes an immunological vulnerability to enteroviruses (the immune system does not attack them efficiently) [4].

The evidence linking rotavirus to type 1 diabetes is based on three pillars: molecular mimicry, animal models and post-vaccination epidemiological data. There is a small protein fragment in rotavirus that shows 56% identity and 100% similarity with an important region of the islet autoantigen IA-2 and another with 75% identity with a portion of GAD65. Both risk regions from rotavirus bind to the HLA-DR4 molecule, which confers susceptibility to type 1 diabetes. In a prospective follow-up study of children with high genetic risk for type 1 diabetes, 86% of anti-IA-2 antibodies, 62% of anti-insulin antibodies and 50% of anti-GAD antibodies appeared or increased further after a rotavirus infection [6].

The epidemiological data regarding the protective effect of rotavirus vaccination on type 1 diabetes incidence are promising. A recent study confirmed a decrease in incidence in 7 out of 8 countries analysed. A recent meta-analysis of 4.4 million children calculated a 13% lower risk in vaccinated children [7].

During the COVID-19 pandemic, cases of hyperglycaemia, diabetic ketoacidosis and newly diagnosed diabetes increased, suggesting that the SARS-CoV-2 virus may be a trigger for type 1 diabetes. Laboratory studies have demonstrated that SARS-CoV-2 can directly infect both pancreatic beta cells (the endocrine pancreas) and the cells responsible for digestive enzyme secretion (the exocrine pancreas). Viral infection reduces insulin secretion capacity and induces the death of some beta cells through their own decision, to protect the others (apoptosis). The incidence of type 1 diabetes was 14% higher in the first pandemic year and 27% higher in the second year. Overall, diabetic ketoacidosis at onset increased by 26%. The risk of triggering the hyperglycaemic form of type 1 diabetes after SARS-CoV-2 infection was 42% higher, rising to +67% in children under 12 years [8].

The indirect factors of the pandemic (reduced access to medical services, diagnostic delays, social isolation) may have been more important than direct viral infection. It appears that there is no significant association between SARS-CoV-2 infection and presymptomatic type 1 diabetes autoimmunity (stages 1 and 2) [1].

The CoviDIAB registry is an international patient registry for new-onset diabetes associated with COVID-19 infection. It was founded by a group of 17 international diabetes experts and announced in the New England Journal of Medicine in June 2020. The registry was created to investigate the bidirectional relationship observed between COVID-19 and diabetes. Diabetes increases COVID-19 severity (20-30% of COVID-19 deaths occurred in people who also had diabetes), and on the other hand there are many cases of newly diagnosed diabetes reported in patients with COVID-19 [9].

The CoviDIAB registry (accessible at covidiab.e-dendrite.com) collects detailed clinical data about patients with confirmed hyperglycaemia, documented COVID-19 infection, no history of diabetes and previously normal glycated haemoglobin (HbA1c) values. The main goals are establishing the extent and phenotype of new-onset diabetes associated with COVID-19, investigating epidemiological characteristics, pathogenesis and obtaining clues regarding the appropriate management of these patients. A fundamental question the registry aims to answer is whether post-COVID diabetes represents classic type 1 diabetes, type 2 diabetes or possibly a new form of diabetes.

Three vaccination strategies are relevant from the perspective of type 1 diabetes prevention: rotavirus vaccination, rubella vaccination and the development of an enterovirus vaccine. Rotavirus vaccination has the best studies regarding the decrease in type 1 diabetes incidence after the introduction of this vaccination [7]. Rubella vaccination (the component of the MMR vaccine) has virtually eliminated congenital rubella syndrome in developed countries and with it, the diabetes associated with congenital rubella.

The most innovative approach is the development of a multivalent, inactivated vaccine targeting Coxsackievirus B 1-5. The scientific rationale is based on evidence that infections with these viruses are environmental factors implicated in the onset of type 1 diabetes. The vaccine is specifically designed to prevent acute infections and thus possibly prevent the autoimmune destruction of beta cells [10].

Although the evidence is less direct than for viruses, gut microbiome alterations and certain bacterial infections can contribute to the onset of type 1 diabetes. There are gut microbial biomarkers in one-year-old infants associated with the future development of type 1 diabetes [11]. It is possible that alteration of the gut flora leads to easier passage of certain toxins from the large intestine into the blood, with as yet unknown effects on the mechanisms related to the onset of type 1 diabetes.

A specific bacterial agent studied is Mycobacterium avium subspecies paratuberculosis (MAP), significantly associated with type 1 diabetes. It appears that there is a very strong association between the presence of anti-MAP antibodies and type 1 diabetes, but no association with type 2 diabetes [12]. Sardinia, which probably has the highest type 1 diabetes incidence in the world (~74/100,000, surpassing Finland), has a particularly high incidence of MAP infection.

The relationship between repeated infections in early childhood and the risk of type 1 diabetes is complex. The hygiene hypothesis suggests that reduced exposure to infections in early childhood (urbanisation, excessive hygiene, antibiotics) deprives the immune system of the training needed to best distinguish between external aggressions and what represents the body's own tissues. This increases the risk of autoimmune diseases. This hypothesis is supported by the higher incidence of type 1 diabetes in countries with higher hygiene standards (e.g. Finland, Sweden) compared with those with lower standards (e.g. Russia, Romania). The accelerator hypothesis focuses on insulin resistance and metabolic stress as factors that accelerate the loss of beta cells initiated through any other mechanism, known or unknown. Repeated infections can contribute through the production of various substances resulting also from the battle with the immune system, which increase insulin resistance [13].

The fertile field hypothesis proposes that inflammation induced by viral infection fertilises the pancreatic environment, thus creating favourable conditions for generating autoreactive T lymphocytes, for a limited period of time. The effect of infections depends on the type, quantity, timing and duration of these infections, as well as the child's genetic background. Prolonged Enterovirus B infections are harmful, while early exposure to adenoviruses can be protective [4].