Type 2 diabetes (T2D) will affect ~650 million adults worldwide by 2040 and about as many will have pre-diabetes. Chronic hyperinsulinemia/insulin resistance precedes T2D development. Studies link insulin resistance with chronic inflammation and oxidative stress.

In Guinea-Bissau, a low-income country in West Africa, the T2D incidence is largely unknown and there is an acute lack of diabetes doctors, nurses and other diabetes educators. They hardly have access to insulin, and mortality from T2D complications is high. Previous studies by the Bandim Health Project (www.bandim.org) in the country show that the Bacillus Calmette-Guérin (BCG) vaccine has non-specific effects, well beyond tuberculosis prevention, conferring a general protection against unrelated pathogens. At the same time, studies from the US have also shown that BCG can significantly improve glycemic control in Type-1 diabetes (T1D) patients. Yet, no such studies have been done in T2D or pre-diabetes.

The purpose of the present study is to administer BCG to patients with pre-diabetes, in order to reduce hyperinsulinemia/chronic inflammation, a novel strategy to flatten the growing T2D incidence.

BACKGROUND The lack of medical infrastructure and health education is clear across Guinea-Bissau, West Africa. Besides pilot studies from the capital Bissau, showing poor glycemic control and high mortality, due mainly to untreated type 2 diabetes (T2D), the diabetes prevalence is largely unknown. Patients cannot easily afford treatment, including insulin. Better health education, cheaper prevention and treatment strategies are needed.

Importantly, data show that the Bacillus Calmette Guerin (BCG) vaccine has key non-specific health effects, well beyond prevention of the target disease tuberculosis. Furthermore, in the US, it has been shown that BCG revaccination lowers haemoglobin-A1c (HbA1c) in people with Type-1 diabetes (T1D). Yet, no such studies have been done in T2D subjects.

The exact pathways of the non-specific effects of BCG are still being exploried, but it may be due to an increase in basal systemic levels of type 1 cytokines and immune cells and induction of epigenetic modifications in the innate immune cells, leading to increased pro-inflammatory responses towards unrelated pathogens. These modifications are accompanied by changes in glucose metabolism, increasing aerobic glycolysis, a state of high glucose utilization. In the randomized trial from the US in T1D patients, BCG lowered HbA1c close to normal levels. A systemic shift in glucose metabolism from oxidative phosphorylation to aerobic glycolysis was observed. BCG vaccination could therefore potentially be a cheap and safe tool to prevent T2D in resource-limited settings.

In Guinea-Bissau, only a couple of diabetes studies have been performed in patients with tuberculosis, HIV, twins and at a local diabetes clinic. Recently, study identified a severe lack of diabetes prevention, management and education. BCG vaccination could thus be a very valuable tool to prevent (and possibly treat) diabetes in low-income countries like Guinea-Bissau.

At the same time, studies indicate that long-term inflammation may predict dysmetabolism, independent of being lean or obese. Chronic inflammation is an important feature of insulin resistance and T2D, in which specific pro-inflammatory cytokines drive disease.

The rationale for the present study in Guinea-Bissau is therefore to test whether BCG vaccination could reduce HbA1c levels and ameliorate inflammation observed in individuals with pre-diabetes, thereby preventing the progression to overt T2D.

AIM OF STUDY Conduct a randomized trial to investigate the metabolic effects of two BCG vaccinations 4 weeks apart in persons with pre-diabetes. Metabolic status is monitored during the trial and related with inflammatory biomarkers.

MAIN HYPOTHESIS Providing two BCG vaccinations 4 weeks apart to individuals with pre-diabetes, compared with placebo, is associated with lower fasting glucose/hba1c levels at 3 years follow-up.

DEFINITIONS Pre-diabetes: F-glucose between 5.6-6.9 mmol/L or HbA1c between 39-47mmol/L. Above these levels is diabetes.

INVESTIGATIONS AND SAMPLE SIZE

The aim is to conduct a randomized trial to investigate the metabolic effects of two BCG vaccinations 4 weeks apart in pre-diabetes adult subjects. The pre-diabetes patients will partly be identified in a diabetes prevalence study being conducted prior to the trial.

The participants are followed for 3 consecutive years, with F-glucose and hba1c measured every 3 months.

Metabolic status is monitored by measuring metabolic markers (glucose, hba1c, insulin, leptin, adiponectin) and relating these with inflammatory markers (Il-6, Il-10, Il-33, TNFa) pre and post-trial.

The blood for the advanced metabolic and inflammatory biomarkers will be collected and stored in frozen condition (-40C) in Guinea-Bissau at the National Health Laboratory (LNSP), before being transported to Coimbra University, Portugal, for analyses.

Sample size: If anticipating at 3 years follow-up a reduction in mean F-glucose from 7.0 mmol/L to 6.0, there would be a need 63 participants in each group, with a potency of 80% and an alpha of 0.05. Therefore, including a total of 200 participants with pre-diabetes in this BCG study should give us an appropriate sample size, also taking into account loss to follow-up.

SETTING The trial will be conducted by the Bandim Health Project (BHP) team in Bissau (www.bandim.org). Since 1978, the BHP has maintained a health and demographic surveillance system (HDSS) site here.

The National Diabetes Clinic (NDC): Is the only diabetes clinic in Guinea-Bissau, situated within the BHP study area, where the trial will take place. The clinic is semi-private and run by the Guinean Diabetes Association (ANDD).

METHODS The investigators will invite 200 individuals with pre-diabetes. Provided the participants meets the inclusion/exclusion criteria, fasting blood samples will be collected by a team nurse. Clinical records will contain demographic characteristics, patient history, anthropometry, clinical measurements, prescriptions, diagnoses and the presence of diabetic wounds. BCG and smallpox vaccination scar status are assessed.

Follow-up: Participants will be recalled for evaluation every three months, during the 3 years trial, where the investigators will measure F-glucose and HbA1c levels.

Glycemic analyses: F-glucose levels will be measured using Accu-Chek Aviva. Hb1Ac will be measured on a Bayer's A1CNow.

Bio markers: At baseline and every 6 months during follow-up the investigators will measure metabolic (glucose, hba1c, insulin, leptin, adiponectin) and inflammatory markers (Il-6, Il-10, Il-33, TNFa).

Data management: After randomization, the participants will receive a card with a unique study number, which they need to bring to facilitate identification. All information will be entered in RedCap.

STATISTICS BCG's effect on F-glucose and hba1c will be done using Poisson regression models, and time to T2D diagnosis by Cox regression. Analyses will be done overall, by sex and pre-diabetes levels. It will be examined whether prior BCG-scaring affects the intervention.

ETHICAL CONSIDERATIONS The protocol has been approved by to the Guinean Ethical Committees. Informed written consent will be obtained before enrolment.

DATA SAFETY AND MONITORING BOARD This will be established consisting of a diabetes expert and a statistician plus an additional member.

NOVELTY AND SIGNIFICANCE Better diabetes prevention and treatment strategies are urgently needed. If the trial is successful, it will represent a major improvement in the diabetes control, and at very low cost. Essentially, with the proper handling techniques, BCG can be administered everywhere, even in the most resource-limited areas.

SPONSOR: The BHP and the University of Coimbra (UC) are both the sponsors of the study, and wrote the protocol jointly. The UC is therefore also a co-responsible party.

Inclusion Criteria for BCG trial:

• Ages 18-64
• Planning to continue living in the study area
• F-glucose from 5.6-6.9 mmol/L and HbA1c between 39-47mmol/mol.

Exclusion Criteria for BCG trial:

• HIV infection (an HIV test to be done before enrolment)
• Pregnancy (a pregnancy test to be done before enrolment in women in the childbearing age)
• Chemotherapy

Deidentified participant data can be shared after approval of a data-sharing proposal submitted to the study chair person (Eugenia Carvalho).

Participants will be blinded to treatment. The investigators (or delegated vaccinator, e.g. nurse) administering the BCG vaccine or placebo (saline) will not be blinded. In case of serious adverse events, the participant can be unblinded after consultation with the coordinating PI.