Developing a Controlled Human Infection Model for Group B Streptococcus (CHIM_GBS)

The global burden of Group B Streptococcus (GBS) is high and represents an unmet public health need. GBS is the leading cause of neonatal sepsis and meningitis in most countries. In 2015, it was estimated that worldwide there were at least 320,000 infants with invasive GBS disease, 90,000 infant deaths and 10,000 cases of children with disability related to GBS meningitis1.

Maternal rectovaginal colonization with GBS is a prerequisite for early onset disease within the first 6 days of life, with vertical GBS transmission occurring around the time of birth2. An estimated 20-35% of pregnant women are colonised with GBS with vertical transmission occurring in approximately 50% of colonized cases3. 1-2% of neonates born to GBS-colonised women develop invasive disease in the absence of intrapartum antibiotic prophylaxis (IAP)4.

Whilst the incidence of Early Onset Disease in the USA has declined significantly following the adoption of a universal swab-based screening and IAP policy, in countries adopting a risk-based IAP strategy, such as the UK, increases in Early Onset disease burden are reported.4. IAP has limitations as a strategy for use in countries such as Uganda due to cost and access to healthcare, as well as the risk of antimicrobial resistance developing.

Given the early onset of neonatal GBS disease and the shortcomings of IAP policies, one of the best approaches to prevent GBS disease could be to use vaccination to prevent colonisation of pregnant women. The ideal model to test any GBS vaccine would be a controlled human infection model in non-pregnant women, however in order to develop this controlled human infection model researchers need to understand how host immunity affects GBS colonisation. Around 25-30% of non-pregnant women are colonised with GBS at any time and colonisation may be lost or gained over a lifetime. Prior to performing controlled human infection models, researchers need to further understand the correlation between colonisation and GBS antibody production in serum and at mucosal surfaces and researchers need to better understand how colonisation changes over time in healthy, non-pregnant women and given the observed differences in GBS disease in high and low income countries it would be useful to perform these observations in different geographical locations.

RATIONALE

Firstly, investigators need to determine levels of pre-existing local and systemic immunity in GBS colonised and uncolonised women that could be used as inclusion/exclusion criteria in a controlled human infection model, and to achieve this investigators need to optimise sampling techniques.

Secondly, investigators need to understand differences in vaginal immunity that may be due to genetic or geographical differences in order to demonstrate the potential for global applicability of any potential vaccine candidate this is the rationale for conducting the study in the UK and Uganda. Finally investigators need to clarify who could (and would) be likely to volunteer to take part in controlled human infection model studies for GBS.

THEORETICAL FRAMEWORK

Correlation between vaginal sero-specific Immunoglobulin G to GBS has been found in pregnant women with less colonisation in those women with higher levels of Immunoglobulin G suggesting a level of natural immunity5. It is unclear how long Immunoglobulin G in blood remains high following colonisation and whether this decreases over time leading to increased future susceptibility. There is little published research on how GBS colonisation naturally changes over time and whether this correlates with vaginal and serum serotype-specific GBS Immunoglobulin G levels.

In order to develop a human infection model, it would be useful to know how natural colonisation varies over time and its relationship with natural immunity, such that for future vaccine studies researchers would know whether observed non-colonisation is due to vaccine or natural immunity.

A secondary aim is to explore whether human infection models of GBS would be acceptable to women given the complex nature of GBS colonisation and its potential impact on subsequent pregnancies.

This will be a prospective cohort study of 100 women (50 colonised with GBS at baseline and 50 uncolonised with GBS at baseline) at each site sampled over 12 weeks to assess antibodies in the vagina every two weeks and in blood every 6 weeks. The study will also include a focus group to understand acceptability of methods for the proposed human infection challenge model of GBS.

1. Screening for GBS Women will be recruited for screening for GBS in the UK and Uganda (250 at each site). At screening women will be consented for a vaginal and rectal swab to assess for GBS carriage and will also undergo an asymptomatic Sexually transmitted infections screen according to local usual practice. In both the UK and Uganda this screening will look for Chlamydia, Gonorrhoea, Trichomonas vaginalis, Bacterial Vaginosis, Candida and for Syphilis, HIV and Hepatitis B and C. Anyone who meets the full inclusion criteria following screening will be invited to take part in the sampling study until the recruitment targets are reached. If any woman tests positive for any of the infections, she will be referred to a local centre for treatment and may still be included after completed treatment for the infection. Of the 250 women screened, the investigators will aim to recruit 50 women who are positive for GBS and 50 negative for GBS to take part in the sampling study. Consent to undertake the screening will be done initially with a further consent process for the sampling study. Screening will involve three self-taken vaginal swabs and a rectal swab for GBS detection and Sexually transmitted infections detection and a venous blood sample for HIV, syphilis and Hepatitis B and C. A brief medical history will be conducted at screening to exclude women with a history of diabetes, genital dermatoses, Cervical Intra-epithelial Neoplasia or other condition which the investigator deems may make them unsuitable to take part in the sampling study. A bedside urine pregnancy test will be performed at screening. HIV and Hepatitis B and C are not exclusion criteria however this information is needed in the analyses as there is a possibility that women with HIV in particular do not have the same antibody response to colonisation.
2. Sampling method optimisation If a woman is deemed to meet all inclusion criteria and no exclusion criteria at screening and is willing to take part in the sampling method optimisation study, she will be consented again for the further study including consent for participation in a focus group at the end of the study. Consent to participate in the focus group will be optional. The sampling study will last for 12 weeks and samples will be collected as follows:
3. A self-taken low vaginal swab using swabs at enrolment and then two weekly intervals, (Day 0, day 14, day 28, Day 42, day 56, day 70 and day 84) for GBS carriage.
4. A self-taken rectal swab at recruitment and then at two weekly intervals (Day 0, day 14, day 28, Day 42, day 56, day 70 and day 84.) which will be tested for GBS carriage.
5. Menstrual cup fluid at recruitment and then at two weekly intervals (Day 0, day 14, day 28, Day 42, day 56, day 70 and day 84) which is tested for Immunoglobulin G antibody (total and GBS specific).
6. Serum sample at day 0 and at day 42 and day 84 for Immunoglobulin G detection (total and GBS specific) - 10 mls of blood will be collected using sterile needle and syringe into serum separator tubes.
7. Urine pregnancy tests will be performed at day 0, day 42 and day 84 to ensure no pregnancies during the study. Women may choose whether to self-sample with rectovaginal swabs and menstrual cups at home or on attendance for their study visit. Blood tests and urine pregnancy tests will always be performed by the study team.
8. Focus groups In the UK the investigator will recruit up to 20 women from the sampling study to take part in focus groups about their experiences with self-sampling methods and the acceptability of controlled human infection models. Consent to participate in focus group discussions will be included as part of the consent to take part in the sampling study but will be optional. Women may opt out of the focus groups at any time but remain in the sampling study if they choose.

In Uganda, the investigators will recruit more widely to our focus groups, including representation from midwives and the participant's partners and community leaders. Investigators will directly contact potential focus group participants. Focus groups will be conducted on the acceptability of controlled human infection models. Investigators will also explore potential issues around maternal vaccination, and traditional and contemporary views on taking vaginal swabs and blood samples. From the Ugandan focus groups the investigator will ask 30 volunteers (10 men and/or community leaders, 10 women and 10 midwives) to participate in three further discussion groups to explore more in-depth views about the safety and acceptability of vaginal swabs and menstrual cups, potential hesitancy in consenting to have samples taken and potential barriers to joining the study.

Results from the focus groups will be used to inform patient information leaflets and sampling protocol for the design of a controlled human infection model study.