Background

Type 1 diabetes (T1D) mellitus is a chronic autoimmune disease characterised by the destruction of insulin-producing β-cells in the pancreatic islets resulting from a complex immunological process that activates both the innate and adaptive immune systems. β-cell-specific effector T cells mediate pancreatic islet inflammation and enhance infiltration of other immune cells such as antigen-presenting cells, macrophages and B cells. These immune cells contribute to a chronic state of islet inflammation through the production of proinflammatory cytokines and pathogenic interactions and pathways (figure 1).1 Interventions targeting proinflammatory cytokines, such as interleukin (IL)-6, IL-8 and tumour necrosis factor (TNF)-α, through antagonism and anti-inflammatory cytokines, such as IL-2 and IL-10, through agonism have been identified as potential therapeutic strategies in T1D.2 Interestingly, studies of cytokine responses after influenza vaccination have shown an upregulation of IL-2 and IL-10, while proinflammatory cytokines such as IL-6, IL-8 and TNF-α are downregulated.3–5 This suggests that influenza vaccination could potentially downregulate the proinflammatory cytokines involved in T1D pathogenesis, additionally upregulate the anti-inflammatory cytokines and reverse or halt the immunopathogenesis of T1D (figure 1).

Figure 1

Immunomodulation of type 1 diabetes by influenza vaccination is shown schematically in the green box. Trained innate immunity is induced and affects innate immune cells such as dendritic cells, macrophages, monocytes and NK cells by triggering metabolic reprogramming and epigenetic modifications. This favours the production of anti-inflammatory cytokines and reduces the proinflammatory cytokine response, resulting in decreased pancreatic islet inflammatory activity and increased β-cell function. The red box illustrates the adaptive immune response involving B and T cells. Cytotoxic effector T cells are directly implicated in the destruction of β-cells in the pancreatic islets. The outcome of the T cell-mediated destruction is an intensified pro-inflammatory response, leading to persistent pancreatic islet inflammation. NK, natural killer.

The rising global incidence of T1D in children and adolescents6 has intensified the focus on preserving insulin-producing β-cells via immunomodulation. The primary goal of disease-modifying therapies is to modulate the autoimmune process and prevent further β-cell destruction and complications related to T1D.7 Preserving residual β-cell function is desirable as it reduces short-term and long-term complications. In the short term, sustained β-cell function provides better glycaemic control, lowers the risk of hypoglycaemia, reduces insulin requirements and decreases the risk of hospitalisation.8 9 Long-term benefits include reduced rates of microvascular complications such as nephropathy, neuropathy and retinopathy.10–12 Influenza vaccination is a safe intervention and is recommended by the WHO for both children and adults with T1D.13 It is uncertain whether the timing of influenza vaccination at the onset of T1D may affect the efficacy of influenza disease. While originally created to safeguard against influenza infection, the influenza vaccination is now being explored for its potential pleiotropic effects.14 15 To our knowledge, the use of influenza vaccination as a disease-modifying strategy to alleviate T1D in children and adolescents has not been previously investigated. Until now, few disease-modifying therapies for patients with recent-onset T1D have shown an acceptable balance between benefits and side effects in sustaining long-lasting residual β-cell function in either children or adults.16 17

In the present study, we hypothesise that influenza vaccination, through the induction of pleiotropic effects, will decelerate immune-mediated β-cell destruction. This hypothesis will be evaluated by measuring 2-hour mixed meal tolerance test (MMTT)-stimulated C-peptide levels following influenza vaccination as compared with placebo to assess preserved β-cell function in children and adolescents with recent-onset T1D.

MethodsStudy design, primary and secondary endpoints

The INfluenza VaccInation To mitigate typE 1 Diabetes (INVITED) trial is a prospective, randomised, double-blind, placebo-controlled, multicentre clinical trial conducted in paediatric and adolescent patients with recent-onset T1D at Danish paediatric departments.

The primary outcome is the difference in mean change from baseline to 12 months in C-peptide level between groups in a 2-hour MMTT. Secondary endpoints are: (1) difference in mean change from baseline to 6 months in C-peptide level in a 2-hour MMTT between study groups, (2) change in haemoglobin A1c (HbA1c) from baseline to 6 and 12 months, (3) change in time-in-range from 6 months to 12 months and (4) change in insulin requirements (measured as average total insulin dose per kg body weight per day over the last 14 days) from baseline to 6 and 12 months.

Exploratory endpoints are changes in diabetes-related autoantibodies (glutamic acid decarboxylase-65 antibodies, insulin antibodies, zinc-transporter-8 antibodies and islet cell autoantibodies) and changes in inflammatory markers (IL-2, IL-6, IL-8, IL-10 and TNF-α) from baseline to 6 and 12 months. In addition, we will measure serum haemagglutinin inhibition antibody titres against the four viruses included in the influenza vaccine at baseline and after 6 and 12 months. The number of unplanned hospitalisations related to diabetes will be reported for the two groups. Primary, secondary and explorative endpoints are shown in box 1.

Box 1 Primary, secondary and exploratory endpointsSecondary endpoints

Change in fasting C-peptide levels, baseline to 6 months; change in HbA1c, baseline to 6 and 12 months; change in insulin requirements, baseline to 6 and 12 months; change of time-in-range, time-above-range and time-below-range, coefficient of variance from baseline to 6 and 12 months.

Exploratory endpoints

Change in diabetes autoantibodies glutamic acid decarboxylase-65 antibodies, zinc-transporter-8 antibodies, islet cell antbodies and insulin antibodies (GAD-65, ZnT8, ICA and IAA), baseline to 6 and 12 months; change in inflammatory markers (IL2, IL6, IL8, IL19 and TNF-α), baseline to 6 and 12 months; serum haemagglutinin inhibition antibody titre against the four influenza viruses in the vaccine, baseline to 6 and 12 months.

Study population and patient selection

Children and adolescents with recent-onset T1D as defined by classic clinical symptoms and random blood glucose >11.1 mmol/L will be invited to participate in the trial.18 Participants will be recruited during hospitalisation at Aarhus University Hospital, Aalborg University Hospital, Regional Hospital Randers, Regional Hospital Viborg, Gødstrup Hospital, Herlev Hospital, Holbaek Hospital, Zealand University Hospital Roskilde, Slagelse Hospital and Zealand University Hospital Nyköbing F., all in Denmark. Candidate participants fulfilling all inclusion and exclusion criteria will receive both written and oral information about the study and be enrolled upon informed consent from a study investigator. Written informed consent will be obtained from both parents (or legal guardians). Study inclusion and exclusion criteria are listed in box 2.

Box 2 Inclusion and exclusion criteriaInclusion criteria

Patients hospitalised with newly diagnosed type 1 diabetes mellitus.

Male or female participants aged ≥7 and <18 years.

Written informed consent (parents or legal guardians).

Exclusion criteria

Influenza vaccination during the current influenza season.

Indications for influenza vaccination due to other conditions.

Severe allergy to eggs or a previous allergic reaction to the influenza vaccine.

Suspicion of a febrile illness or an acute, ongoing infection.

Patients with endogenic or iatrogenic immunosuppression that may result in reducing immunisation response.

Inability to provide informed consent from a parent or legal guardian.

Randomisation

Randomisation is performed in a 1:1 fashion, stratified by age groups (from 7 to 12 years of age or from 13 to 17 years of age) and sex (figure 2). Randomisation will be performed in a study-specific online electronic case report form (eCRF) provided by Research Electronic Data Capture hosted at Aarhus University.19 20 Randomisation will occur after the participant has been enrolled.

Figure 2

Consolidated Standards of Reporting Trials flowchart.

At each study site, an unblinded study nurse, not otherwise involved in or participating in the study, will perform the randomisation in the eCRF and prepare syringes for administration of either vaccine or placebo. To ascertain blinding, the nurse wraps a piece of foil around the syringe to ensure that the patient or parents cannot see what is administered during the injection. The unblinded study nurse will keep a randomisation log with the identification numbers of the participants and information on the administration of influenza vaccination or placebo. The randomisation log will be securely stored, accessed only by the unblinded study nurse and collected at the end of the trial. Unblinding is permissible solely for participant safety in the event of a serious adverse event. Participants, investigators and all medical staff will be blinded to allocation until the last study, follow-up of the last patient and statistical analysis have been performed. All study participants will receive standard medical care. The regime for enrolment, data collection, interventions and assessments of the trial are shown in figure 3 and the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) reporting guidelines have been used.21

Figure 3

Standard Protocol Items: Recommendations for Interventional Trials figure of the trial procedure, including enrolment, data collection, intervention and assessments.

Influenza vaccine

Seasonal inactivated standard-dose quadrivalent influenza vaccines according to WHO recommendations (VaxigripTetra, Sanofi, 0.5 mL) will be used.13 Recruitment was started in December 2022, and we plan to enrol patients during three influenza seasons, extending into the non-influenza season until the expiration of vaccines. During the baseline visit, participants will receive an injection of either the influenza vaccine or placebo (isotonic sodium chloride, 0.5 mL). All additional vaccinations during the study period will be recorded at follow-up visits. Participants are allowed to receive influenza vaccination during the study period as per the recommendation to patients with poorly controlled T1D.22 23

Clinical examinations and tests

Participants will be assessed at baseline during hospitalisation (or a maximum of 14 days after diagnosis) and at 6-month and 12-month follow-ups in a research unit or outpatient clinic. Examinations and MMTTs will be conducted, and samples of biological material will be collected at all three study visits.

Examination

Heart rate and blood pressure will be measured following a 15-min rest by a digital automatic sphygmomanometer. Height and weight will be measured to obtain the body mass index. Skinfold measurements at four points—triceps, biceps, subscapular and suprailiac—will be conducted to calculate lean body mass and fat percentage. To assess pubertal development, the Tanner scale will be used, ranging from prepubertal (stage 1) to adult phenotype (stage 5).24

Mixed-meal tolerance test

The MMTT is the gold standard test to determine residual β-cell function by measurement of stimulated C-peptide.25 26 C-peptide, a connecting peptide of insulin synthesis, is produced in equimolar amounts as endogenous insulin. C-peptide is not a product of therapeutically administered exogenous insulin and can assess endogenous insulin secretion.27 If a participant normally receives long-acting insulin in the morning, this will be administered after the MMTT. Short-acting insulin can be administered up to 2 hours before the start of the MMTT. Participants are required to fast from 22:00 the evening before the visit. Fasting is terminated by the end of the MMTT. If glucose-holding drinks or foods are ingested due to hypoglycaemia during the fasting period, the next-day MMTT will be cancelled and rescheduled. Blood glucose is aimed at a normal level before MMTT and should increase to maximal stimulation of endogenous insulin secretion. Blood samples will be collected before ingestion of the standard MMTT drink at −10 min and after ingestion at 15, 30, 60, 90 and 120 min. At the end of the MMTT, blood glucose will be normalised with the required amount of insulin to normalise blood glucose.

Blood and urine samples

Venous blood samples will be collected at baseline and during follow-up visits at 6 and 12 months. Blood samples will be centrifuged immediately and stored at −80°C pending further analyses. All samples will be analysed in batches.

Urine samples will be collected at baseline and follow-up, before the MMTT at −10 min and after the MMTT at 120 min. All samples will be stored at −80°C pending further analyses. Urine will be analysed for C-peptide concentration and creatinine to estimate a urine C-peptide–creatinine ratio.

Side effects questionnaire

A questionnaire concerning the side effects of the influenza vaccine or placebo will be sent to the participant or parents 7 days after the baseline visit. The questionnaire contains questions regarding local and systemic events during the last 7 days. The safety of influenza vaccination in recent-onset T1D will be evaluated.

Insulin requirements and glucose monitoring

At each study visit, the mode of insulin delivery, basal insulin requirements, measured in units per day and estimated mean bolus insulin requirements during the last 14 days (or a shorter average period at the baseline visit) will be recorded. All study participants will have a sensor for continuous glucose measurements applied during the first admission. Time-in-range (%), time-above-range (%), time-below-range (%) and coefficient of variance (%) will be obtained from the sensor report at follow-up visits at 6 and 12 months.

Data management, sample size calculation and statistical analysis

All trial data from visits 1–3, the 7-day side effects questionnaire and suspected unexpected serious adverse reactions will be recorded in the eCRF. Data management and statistical analyses will be performed at Aarhus University Hospital. Access to the trial database is password-protected and will be restricted to certain study team members with different levels of access assigned by the study coordinator. All participants will be assigned an identification number, recorded in the eCRF, along with baseline data and all data collected within the study.

The study results will be analysed according to the intention-to-treat principle, with one modification. Due to the study intervention (influenza vaccine or placebo) being administered only once during the initial hospitalisation and because most protocol deviations related to the intervention are expected to be logistical (such as patients leaving the hospital before receiving the study medication), we will define the modified intention-to-treat population as all patients who were randomised to their assigned treatment and who received the study medication. A participant can withdraw from the study at any time. Participants who are lost to follow-up at 6 months will be censored on the day of the baseline visit and participants who are lost to follow-up at 12 months will be censored at the 6-month follow-up visit.

Sample size calculation

The sample size calculation is based on recommendations from an American Diabetes Association workshop report25 and a recent publication addressing the opportunity to use analysis of covariance to improve precision.28 The increase in precision is due to adjusting for both baseline C-peptide levels and age, which allows a 50% reduction in sample size relative to the unadjusted comparison. Following recommendations and calculations from Krogvold et al,17 we assumed an effect size of 50% improvement in stimulated C-peptide level at 12 months in the treatment group compared with the placebo group. This assumption is based on considerations of clinical relevance in children and adolescents with T1D. Based on the above-mentioned studies, the sample size calculation was set to estimate a 50% improvement in the primary endpoint in the treatment group with a two-sided alpha of 0.05. Accounting for 10% dropout, we will randomly assign 50 participants in each of the two groups to obtain a statistical power of 85%.

Statistical analysis

The primary endpoint of the relative change in mean 2-hour C-peptide area under the curve (AUC) from baseline to 12 months will be analysed using a linear mixed model for repeated measurements with adjustments for sex, age and baseline C-peptide response in the two study groups. C-peptide AUC will be fitted on a log-transformed scale. In the model, we will include measurements taken at 6 months, as these will improve precision in estimates of individual random effects, which in turn will improve the precision of effect estimates. Differences at baseline between study groups will be assessed with an unpaired t-test on the original scale or a log-transformed scale based on the evaluation of a normal distribution. Non-parametric comparisons will be applied for outcomes where it is not possible to obtain an adequate normal distribution, with or without logarithmic transformation. Ordinal variables will be assessed with the χ2 test for trend or the Mann-Whitney U test. To test for differences between proportions, Pearson’s χ2 test or Fisher’s exact test will be used. Missing data will be assumed to be missing at random. Multiple imputations will be used for individuals if more than 5% of the covariates are missing (m=100). Two-sided statistically significant levels of 5% will be used and estimates will be presented with 95% CIs.

Trial administration

The steering committee consists of investigator and study coordinator IBP, principal investigator OF, KK and MK. Newsletters including recruitment status and contact with site investigators will ensure adequate participant enrolment to reach the target sample size. Additional sites may be added during the study period to ensure the enrolment of 100 participants.

Study monitoring and data safety monitoring

The study will be monitored by the Good Clinical Practice unit at Aarhus University Hospital and Aalborg University Hospital. Any side effects and/or serious adverse events will be recorded in the eCRF and reported to the sponsor. During the study period, the study coordinator will have regular telephone contact with the participating sites to ensure compliance with the study protocol.

An independent data safety monitoring board will monitor study endpoints through an interim safety analysis for a maximum of 1 month following the inclusion of the first 30 participants. Variables to be assessed are key safety outcomes, clinical events, adverse and serious adverse events and study medication side effects as reported in the 7-day patient questionnaire.

Ethics and dissemination

The INVITED trial has been approved by the European Medicines Agency Clinical Trials Information System (CTIS) on 15 September 2022 and has EudraCT number: 2022-500906-17-01. The study is registered at ClinicalTrials.gov (ID NCT05585983). The approval is assessed by the Danish Medicine Agency and the Medical Ethical Committee in Denmark by CTIS. The dissemination of the study results will occur via peer-reviewed publications and conferences.

Current trial status

The current working protocol is V.1.7. Recruitment of participants commenced in December 2022 and the planned end date is December 2024. At present, 32 participants have been enrolled in the trial.

Patients and public involvement

None.

Discussion

To date, no treatments to sustain residual β-cell function in recent-onset T1D have been approved. In the INVITED trial, we will investigate if the administration of influenza vaccination can curb β-cell destruction. The effect will be measured by the stimulated C-peptide response in children and adolescents with recent-onset T1D.

Disease-modifying interventions

Several disease-modifying therapies for T1D have been tested over the past 30 years. Randomised trials testing different agents to modulate the immune system, by inducing T cells, B cells or cytokine responses, have been performed.29 30 However, only a few show promise to sustain residual β-cell function in recent-onset T1D (also designated stage 3 diabetes). One promising disease-modifying treatment is teplizumab, an anti-CD3 monoclonal antibody. Recently, in a randomised trial, teplizumab administered to patients with recent-onset T1D through two 12-day courses of intravenous treatment showed a 59.7% lower decrease in C-peptide response from baseline to 78 weeks between the two groups, in favour of teplizumab treatment.16 Currently, teplizumab is US Food and Drug Administration- approved to delay the progression from stage 2 to stage 3 diabetes, but not for patients with recent-onset T1D (stage 3). Other randomised trials, using rituximab (anti-CD20 monoclonal antibodies) and abatacept (co-stimulation modulation, anti-CD80/CD86 and fusion protein), have also shown significant results in preserving residual β-cell function in children with recent onset T1D, where the rates of decline of C-peptide were shifted by 8.2 months and 9.6 months, respectively.31–33 While adverse reactions to immune therapies are generally transient and self-resolving, adverse events such as neutropenia, infections and cytokine release syndrome may necessitate the discontinuation of treatment.31–34

Different types of vaccines have been explored for their potential immunomodulating, non-specific effects. Multiple studies have focused on the BCG vaccination, which has shown divergent results in T1D depending on vaccination dosing, timing and strains.35–38 Some studies show promising results in long-term improvement in glycaemic control after BCG vaccination, while others find no effect in glycaemic control or improvement in stimulated C-peptide levels.37 38

Influenza vaccination and T1D

The proposed immunomodulating effects of vaccines suggest potential benefits for the immune system extending beyond protection against the specific target disease. Influenza vaccination has been suggested to have such pleiotropic effects on modulating the immune system. Promising but mixed results concerning influenza vaccination in children genetically predisposed to T1D were found in The Environmental Determinants of Diabetes in the Young study.39 Pandemrix influenza vaccination against ‘swine influenza’ showed that Finnish children receiving influenza vaccination had a lower risk of islet autoimmunity and a longer time to diagnosis of T1D compared with non-vaccinated children. A similar effect was not seen in Swedish children, and the risk of type 1 diabetes was unchanged independently of vaccination status.39 In a large Norwegian cohort study, with 2.5 million individuals below 30 years of age, a twofold higher risk of T1D was observed in individuals with laboratory-confirmed influenza A (H1N1) infection.40

Pleiotropic effects and trained immunity

The pleiotropic effects induced by vaccines have been investigated for several years, especially with regard to child morbidity and mortality.41–43 The precise mechanisms underlying these effects remain unclear, but activation of trained innate immunity and enhancement of adaptive immune cells may be possible explanations.44 Trained immunity is a process in the innate immune cells that exhibits adaptive characteristics in memory. These immune cells provide immune memory through epigenetic reprogramming of transcriptional pathways and metabolic reprogramming (figure 1). It is suggested that trained immunity may hold broader implications not only for host defence against infection but also for chronic inflammatory diseases, such as T1D.44 45 In a randomised trial investigating the effects of influenza vaccination in patients with acute myocardial infarction, a significant protective effect on mortality and morbidity after 1 year was shown. Administration of the vaccine was safe during the acute inflammatory phase after myocardial infarction, and the effect was larger than would be expected from protection against influenza illness alone and extended into the non-influenza season, suggesting an immunomodulating effect.46

In an observational and ex-vivo study, Debisarun et al demonstrated the protective pleiotropic effects of influenza vaccination in relation to the incidence of COVID-19 infection.14 These effects are manifested through a reduction in systematic inflammation and long-term transcriptional reprogramming of immune cells. By stimulation of peripheral blood mononuclear cells with heterologous ligands, the trained immunity programme was induced, characterised by lower production of proinflammatory cytokines and higher secretion of anti-inflammatory cytokines.14

We hypothesise that influenza vaccination may not merely suppress but actively engage the immune system to modulate the immunopathogenesis of T1D. The immune system, upon recognising a foreign antigen introduced by the vaccine, could redirect its focus and alleviate the immunological ‘pressure’ on β-cells.47 The potential protective immunological effect of the influenza vaccine is based on a redirection of the proinflammatory cells and cytokines theory.46 This theory regarding inflammation may be comparable to recent-onset T1D, which is dominated by extensive inflammation in the pancreatic islets.48

Limitations

Immunomodulation may be more efficacious and important in younger children due to the more aggressive progression of β-cell destruction.29 Previous research indicates that children below 7 years of age have a higher infiltration of CD8+T cells, CD4+T cells and macrophages at the time of diagnosis compared with children in their mid-teens.29 In our study, children under 7 years of age are not included. The Danish national childhood vaccination programme has since September 2022 recommended vaccination of children from 2 to 6 years of age with live nasal influenza vaccination (Fluenz Tetra).49 It would therefore be unethical to include children in this age group due to placebo. An additional limitation pertains to the absence of participant immunological status documentation at the time of randomisation. It is anticipated that the pleiotropic effects of the vaccine may not be severely influenced by the current immunological status, given that the primary mechanism involves the attenuation of the inflammatory response.3 46

Perspectives

If this study demonstrates positive results in preserving residual β-cell function, influenza vaccination could be readily incorporated into the standard care of paediatric patients with T1D both in low- and high-income countries, as influenza vaccination is an inexpensive intervention. To our knowledge, the potential protective pleiotropic effects of influenza vaccination in children and adolescents with T1D have not been previously investigated. Given the increasing focus on the pleiotropic effects of vaccines, further exploration in this area is both highly relevant and warranted.50 Even small or modest protective effects of influenza vaccination on β-cell function could suffice to show a proof-of-principle of immunomodulation in T1D and open the field for further research.