Developing a serocorrelate of protection against invasive group B streptococcus disease in pregnant women: a feasibility study

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 17/153/01. The contractual start date was in April 2018. The draft report began editorial review in February 2019 and was accepted for publication in August 2019. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Kirsty Le Doare has received a travel grant from Pfizer (New York, NY, USA). Paul Heath is an investigator for clinical trials (not group B streptococcus-specific) who carries out work on behalf of St George’s, University of London, and is sponsored by various vaccine manufacturers, including Novartis (Basel, Switzerland), Pfizer and GlaxoSmithKline plc (Brentford, UK). Asma Khalil has been a member of the Health Technology Assessment General Committee since November 2018, and is due to be active until 2022.

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Carreras-Abad et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2019 Queen’s Printer and Controller of HMSO

The global burden of group B streptococcus is therefore high and represents an unmet public health need

Intrapartum antibiotic prophylaxis can reduce the incidence of EOD, and many high-income countries have established IAP policies. The incidence of EOD in the USA has declined significantly in the era of IAP and the incidence is also generally declining in other countries adopting a swab-based screening policy. 7,8 However, in some countries, particularly those adopting a risk-based IAP strategy, such as the Netherlands9 and the UK,10 recent increases in disease burden have been reported.

It is clear that even strict and universal implementation of IAP guidelines does not eliminate EOD,11,12 as disease can occur because of limitations to IAP administration (e.g. in precipitate labour), in infants of mothers who were negative on screening and where no risk factors are evident in labour. 13 Most significantly, IAP has no impact on GBS-related prematurity or stillbirths or LOD, where the burden of disease is substantial. The majority of GBS meningitis occurs after the first week of life, so this particular burden remains. In the USA and the UK, GBS is now the most common cause of bacterial meningitis in children aged < 5 years. 14,15

Given the very early onset of neonatal GBS disease, the shortcomings of IAP-based prevention strategies and the evidence that suggests that maternal antibodies acquired after natural exposure (when transmitted transplacentally to the fetus) may protect the young infant from invasive infection,16 the prospect of protecting mothers and their infants through vaccination in pregnancy is an attractive one. Possible candidates for an effective vaccine include one or more of the conserved surface proteins or the capsular polysaccharide (CPS). 17

Multiple studies of CPS–protein conjugate vaccines in non-pregnant and, more recently, in pregnant women have established the immunogenicity and safety of these candidates. 17 Recent estimates suggest that an effective GBS maternal vaccine (> 80% efficacy), with high (90%) global coverage, could prevent 231,000 infant and maternal GBS cases, 41,000 stillbirths and 66,000 infant deaths annually. 6

Several obstacles exist in moving the most advanced vaccines into Phase III clinical trials. The first is that, given the relative rarity of GBS disease in Europe and the USA, large numbers of infants would need to be recruited to determine vaccine efficacy. 18 Second, obstacles exist in determining what concentration of antibody is required to protect the infant for the duration of the at-risk period (i.e. the first 3 months of life), as there are currently no internationally recognised standards with which to interpret individual study results. 1 Licensure and policy decisions would be significantly accelerated if an immune marker, measured in an analytically and clinically validated assay, was established as a correlate of protection (CoP). Licensure in such a scenario would come with a commitment to establish effectiveness post licensure in a Phase IV study. This was the approach used for licensure of meningococcal C and meningococcal B vaccines. 19

Primary objective

To test the feasibility of collecting serum at delivery (maternal, cord or both) from a large cohort of pregnant women.

Secondary objectives

To test the key operational aspects for a proposed large serocorrelates study:

• enrolment rate [the rate (proportion) of eligible women who are willing to participate in the delivery blood collection study]

• maternal and/or cord blood collection rate

• key clinical exclusion data collection rate [weeks’ gestation at birth, receipt of IAP in labour (yes/no), type of IAP (list), time between administration of IAP and delivery (in hours)]

• infant iGBS surveillance consent rate.

In a substudy of the main study above to assess:

• rectovaginal swab consent rate

• rectovaginal swab collection rate

• rectovaginal GBS colonisation rate

• rectovaginal GBS CPS serotype-specific colonisation rates.

In the substudy above, where samples of maternal/cord blood and rectovaginal swabs are all available, to assess:

• infant blood sample consent rate

• infant Guthrie card consent rate

• infant Guthrie card collection rate.

To test the feasibility of collecting samples from iGBS cases:

• maternal blood and rectovaginal swab, and the baby blood consent and collection rate from the participants of the iGBS study

• maternal and baby blood sample consent and collection rate from national surveillance (all NHS trusts in England and Wales).

Exploratory objectives

To assess:

• the impact of timing of processing and blood sample storage conditions on total IgG concentrations

• the serotype-specific GBS anti-CPS IgG concentrations in maternal serum and cord blood in subjects colonised with GBS at delivery

• the correlation between two different culture techniques for detection of GBS in rectovaginal swabs – enrichment culture medium and direct plating using selective agar.

Study design

This was a prospective cohort study of pregnant women and their infants. The study design is shown in Figures 1 and 2.

FIGURE 1.

Flow chart for the feasibility study in five hospitals in London and the South East.

FIGURE 2.

Flow chart for the national surveillance of iGBS. O&G, obstetrics and gynaecology; PHE, Public Health England.

Participants

Ethics approval and research governance

Ethics approval for the study was given by the West Midlands – Solihull Research Ethics Committee (REC) on 15 June 2018 with the reference number 18/WM/0147. The study was also approved by the Health Research Authority (HRA) and Care Research Wales on 15 June 2018. Site-specific capacity and capability was then given at the different research and governance departments of St George’s University Hospitals NHS Foundation Trust, Kingston Hospital NHS Foundation Trust, Poole Hospital NHS Foundation Trust, Surrey and Sussex Healthcare NHS Trust and Croydon Health Services NHS Trust.

The Antenatal and Newborn Research Advisory Committee approved the Guthrie card substudy on 3 October 2018.

During the study period, the protocol was amended three times, as follows.

Recruitment procedure

Patient recruitment at a site commenced after REC, HRA and local approvals and all subjects were screened and consented by delegates of the chief investigator.

Study information was available on social media, posters on notice boards, institutional and GBSS websites [https://gbss.org.uk/recent-research (accessed 30 January 2018)], letters, text messages or through direct approach in antenatal clinics. Women had the opportunity to ask questions of the research team by telephone or e-mail. If a woman was interested in participating, this was recorded on handheld notes to ensure that delivery staff were aware of her interest.

Informed consent

Informed consent from the participant, the legally authorised representative or the parents/guardians/person with legal responsibility for children was obtained following explanation of the aims, methods, benefits and potential hazards of the study.

Information about the study was given by one of the research midwives/clinical team and women had the opportunity to ask questions. There was no minimum period between receiving information and providing consent. Initially, formal written consent for participation in the study was taken at enrolment and then confirmed with the mother verbally at, during or following delivery. In some cases, there was insufficient time to obtain written informed consent from participants who were presenting and being consented in labour, although appropriately sensitised to the study. Following an amendment submitted to and approved by the ethics committee (on 5 July 2018), we amended the protocol so that we were able to obtain retrospective written consent after sample collection during delivery if verbal consent was previously recorded in the participant’s notes. Written consent could be taken retrospectively within 24 hours of birth. No data were collected until written informed consent was obtained and participants could withdraw at any time.

Parents of infants identified as having a positive culture for GBS from a normally sterile site (blood or cerebrospinal fluid) from participating hospitals in England and Wales were approached for consent to obtain blood samples from the mother and baby during the acute admission and for collection of the GBS isolate. Samples were collected when parents had had time to consider the information they had received about the study and had given full written informed consent to the local study team.

A copy of the signed informed consent form along with a copy of the most recently approved patient information sheet were given to the study participant. An original signed and dated consent form was retained in the investigator site file and a third copy was placed in the medical notes.

Infant follow-up

A follow-up telephone call was scheduled at 90 days after the baby’s birth to establish whether or not the infant developed iGBS during this time period. This telephone call was made by the research midwives at each site in order to keep personal data locally.

National invasive group B streptococcal disease surveillance

Microbiologists and paediatricians at all NHS trusts from England and Wales were informed about the study through established research and clinical networks [British Paediatric Allergy, Immunity and Infection Group (BPAIIG), Public Health England, the Neonatal Infection Network (neonIN) and the UK Paediatric Vaccine Group]. Paediatricians and microbiologists were asked to make contact with study staff when a case of iGBS occurred at their hospital. After having set up the site, the relevant paediatrician provided study information to the parents and sought their permission for inclusion in the study. Following consent from the parent, a sample of blood was obtained from the mother and from her baby and processed as above. The participating mother was also asked to provide a single rectovaginal swab, taken either by the midwife or by mother herself, to be processed as above. Furthermore, the study team contacted the microbiology laboratory and asked for the bacterial isolate in order to culture and fully characterise the GBS strain. These samples and the bacterial isolate were sent to and stored at St Georges, University of London, in a linked anonymised fashion. In addition, parents were asked for their permission to collect the Guthrie card of the infant.

Data collection

To standardise recruitment processes across the study sites and maximise data quality, we used Research Electronic Data Capture (REDCap; 8.1.8, Vanderbilt University, Nashville, TN, USA)30 hosted at St George’s, University of London. A range of data validation checks were carried out in both REDCap and Stata^® version 15 (StataCorp LP, College Station, TX, USA)27 to minimise erroneous or missing data.

Statistical design

Objective 1

To test the feasibility of collecting serum at delivery (maternal, cord or both) from a large cohort of pregnant women.

Enrolment rate

Overall, 22% (95% CI 21% to 23%) of all women delivering during the study period consented to take part in the feasibility study. However, when based on the number of women approached, the enrolment rate was 74% (95% CI 71% to 76%) (Table 2). Two sites did not collect information on the number of women approached, as the clinical teams approaching women did not accurately record when women declined to participate.

Rectovaginal group B streptococcus serotype-specific colonisation rates

A total of 115 out of 120 positive swabs were serotyped with a rapid LAT and PCR was performed when the sample was found to be non-typeable by LAT. The most common serotype was serotype III (30%), followed by Ia (25%), II (16%), Ib (16%), V (14%) and IV (1%) (Figure 4).

FIGURE 4.

The GBS serotype distribution in colonised women.

Participants from national invasive group B streptococcal disease surveillance

The protocol was amended on 6 September 2018 to include national (England and Wales) surveillance for iGBS. This resulted in 18 notifications, from which 11 cases were recruited into the study (58%). Seven cases were not recruited as there was insufficient time to set up the site before the patient was discharged. All parents approached consented to participation in the study. The clinical presentation was EOD in nine and LOD in two cases (both of which were preterm infants) (Table 16).

TABLE 16 - The iGBS cases from the national surveillance

ROM, rupture of membranes.

Case	Infant’s sex	Age at presentation	Weeks’ gestation	Duration of ROM	IAP	Time of first dose	Presentation	Isolate serotype
1	Male	18 hours	40+2	< 18 hours	No	–	Sepsis	III
2	Male	14 hours	42	< 18 hours	Cefuroxime + metronidazole	< 2 hours	Sepsis	II
3	Female	2 hours	41	< 18 hours	No	–	Sepsis	Ia
4	Female	16 hours	38+3	> 18 hours	No	–	Sepsis	Ia
5	Female	< 24 hours	37+4	< 18 hours	No	–	Sepsis + meningitis	Unknown
6	Female	6 hours	39	> 18 hours	No	–	Sepsis	Ia
7	Female	7 hours	40	> 18 hours	No	–	Sepsis	III
8	Male	42 days	30+2	> 18 hours	No	–	Sepsis	III
9	Female	3 days	Term	< 18 hours	No	–	Sepsis	IV
10	Male	5 hours	40	> 18 hours	No	–	Sepsis	III
11	Male	53 days	25	< 18 hours	No	–	Sepsis	II

In total, serotyping data were available for 14 of the 15 iGBS cases (including the four cases from the feasibility study); the distribution is shown in Figure 5. None of the infants whose samples were collected died during the study.

FIGURE 5.

Serotype distribution in iGBS cases.

Exploratory objectives

Impact of timing of processing and blood sample storage conditions on immunoglobulin G concentrations

There was an acceptable impact on total IgG concentrations for periods of time from 6 hours up to 1 week between sample collection and spinning time (Figure 6). The average time of spinning and freezing samples varied between hospitals but was generally within 1 week of sample receipt (Table 17). St. George’s Hospital processed all samples from Croydon, Kingston and St. George’s Hospitals.

FIGURE 6.

Total IgG concentrations by time between sample collection and centrifuging of sample. Guthrie refers to IgG detection from Guthrie cards.

TABLE 17 - Description of timing of blood sample processing by hospital

CB, cord blood; MB, maternal blood.

Laboratory	Description	Number of samples	Mean (days)	Standard deviation (days)	Minimum (days)	Maximum (days)	Total (days)
Surrey and Sussex Healthcare NHS Trust	MB collection to spin	23	1.4	1.0	0.0	4.0	23
	MB spin to freeze	23	0.0	0.2	0.0	1.0
	CB collection to spin	388	1.8	1.2	0.0	7.0	389
	CB spin to freeze	389	0.1	0.6	0.0	4.0
St George’s University Hospitals NHS Foundation Trust	MB collection to spin	349	6.5	6.5	0.0	44.0	366
	MB spin to freeze	366	6.4	6.6	0.0	42.0
	CB collection to spin	480	7.0	6.7	0.0	44.0	524
	CB spin to freeze	524	7.3	6.7	0.0	42.0
Poole Hospital NHS Foundation Trust	MB collection to spin	13	4.2	8.1	0.0	31.0	13
	MB spin to freeze	13	0.0	0.0	0.0	0.0
	CB collection to spin	457	1.5	1.8	0.0	33.0	458
	CB spin to freeze	458	0.0	0.4	0.0	8.0

Feedback from sites

We undertook site visits to all of the participating hospitals on two occasions during the study period. Feedback from the local teams resulted in changes to study design, including recruitment strategies and data and sample collection. Details of feedback from sites can be found in Appendix 1.

Limitations and recommendations

In summary, this initial feasibility study has been able to assess key operational aspects of the large study that is needed to define the serocorrelates of protection against iGBS.

The main limitation is that we were not able to recruit the (hypothetical) target of 4000 women, mainly because of the short period of the study and insufficient staff dedicated to it. However, we believe that we have recruited a suitable number to allow us to draw conclusions that will now aid the design and co-ordination of the main trial. This includes the clinical data to be collected, how they are recorded, when to approach women in pregnancy and how the issue of cord blood collection can be managed. For example, we believe that an opt-out approach, whereby cord blood sampling would become part of normal routine practice, would help to recruit the large cohort we are aiming for in the serocorrelates study. We propose to assess this in a continuation of this study.

Sample size calculation

Based on previous correlates of risk reduction studies,8,20,22 correlates that give risk reductions of 80–90% are feasible and acceptable. An 80% reduction in a case–control study implies an odds ratio of 0.2. The precision around 0.2 for varying case sample sizes and varying proportion of controls with IgG levels above the potential correlate levels (‘cuts’) is shown in Appendix 2.

Possible strategies for completing the case–control study

On discussion with the GBS3 trial team, it is apparent that the most efficient way of recruiting the control group would be by extending the feasibility study rather than by recruiting these women within the GBS3 trial. This is because, in the context of the GBS3 trial, the women who would potentially be eligible to form the control group (i.e. women who are having a swab performed) will (essentially) all go on to receive IAP if they are found to be GBS colonised. By definition, if a woman has received IAP then she would be ineligible to be in the control group as the receipt of IAP will alter the risk of EOD in her baby. It is likely that women in the ‘standard of care’ arm in GBS3 will be less likely to receive IAP but, as they are not having a swab performed, they would not be eligible as controls. Based on the data from this feasibility study, in which 8.3% of women at the three sites that did not participate in the colonisation substudy happened to have a positive swab obtained before birth (89/1072, excluding missing data), we would need to recruit 12,000 women to identify the 1000 in the ‘standard of care’ arm who have been opportunistically swabbed and are colonised in order to capture the 380 infants born to mothers colonised with GBS serotype III.

An alternative for generating the control group would be to identify women in GBS3 who have a positive GBS swab result, but (for whatever reason) do not receive IAP. This is not our preferred method, as it seems likely that they will be an unusual (and potentially unrepresentative) group of women. However, more importantly, the GBS3 team have indicated that it will not be possible to identify (at an individual level vs. a unit level) those women who actually receive IAP. Owing to the design of the GBS3 study, it will therefore be impossible to identify these women.

The feasibility study has allowed us to identify best practice in recruiting women to have a rectovaginal swab, cord blood and key clinical information collected. We believe that the control group of 5000 women could therefore be recruited using the five sites that have participated in the feasibility study; all sites have expressed a willingness to do so, subject to sufficient funding to support recruitment.

We have recruited around 600 women with swabs in 4 months from two sites (75 per site per month). We therefore estimate that we will require an additional 12 months at five sites to complete control group recruitment (i.e. 75 women per month per site = 75 × 5 = 375 women per month; 375 women × 12 months = 4500 women), plus 3 months for site set-up (renewal of ethics and local approvals) and to ensure staff coverage within the midwifery teams (total 15 months extension). This would ensure that we achieve our target of 5000 women, of whom we expect 380 to be colonised with GBS serotype III and who would serve as our control group for iGBS III cases, recruited in the main study.

Invasive group B streptococcal disease cases from the proposed main trial (GBS3)

As indicated, cord blood will be required from a cohort of approximately 180,000 women. The feasibility study has indicated that recruitment to a study in which cord blood is to be obtained can be very high when a study midwife has the opportunity to approach a woman to obtain consent (74%, 95% CI 71% to 76%). However, as it was possible to approach only a minority of women during the study period because of the individual consent approach, the overall recruitment rate (22% of all births during the study period, 95% CI 21% to 23%) was much lower. Performing clinical trials for an intrapartum intervention has historically been challenging because of issues related to consent in an ‘emergency’ setting. Advances in experience and knowledge of delivering high-quality and high-impact research in labour have however recently emerged. A consent pathway developed by the GBS3 trial group32 now features in the RCOG guidance on intrapartum consent. 33 Recent examples of successful intrapartum emergency studies include deferred or immediate cord clamping in preterm birth,34 glyceryl trinitrate for retained placenta,35 tranexamic acid for treatment of post-partum haemorrhage,36 with further trials under way [e.g. the High Or Low Dose Syntocinon for delay in labour (HOLDS) trial ISRCTN 99841044 (Kenyon S, Taylor R, Hewston R, Johnston T, Hinshaw K, Middleton L, et al. Birmingham Women's NHS Foundation Trust, March 2016 to June 2019) and ANODE ISRCTN: 1116698437]. Feedback from the participating sites suggests that obtaining cord blood from women in labour using an opt-out strategy would be both realistic and acceptable. We propose testing this strategy more formally through patient groups and by seeking ethics committee advice.

To identify those infants in this cohort in GBS3 who develop iGBS we propose to undertake surveillance via Public Health England and specific paediatric clinical networks (BPAIIG and neonIN). We have successfully identified cases through these networks in the feasibility study and have been able to capture 15 cases of iGBS in a short period of time. This provides confidence in our approach. An additional advantage to extending the feasibility study as proposed above would be to continue to test the feasibility of national surveillance and to embed it as a routine, in preparation for the GBS3 study.

In addition, an extension would allow us to test the hypothesis that antibody concentration at the time of acute disease equates to antibody concentration from cord blood. We have already enrolled 27 babies born to healthy women into our antibody kinetics study (eight colonised with serotype III). However, to ensure that we understand the relationship between antibody in the acute disease phase and in a healthy infant exposed to GBS, we now need to undertake the kinetics study in cases of GBS disease up to 3 months of life. Collecting acute serum from 50 infants with serotype III disease will enable us to interpret antibody in health and disease up to 90 days of life. If infant antibody at the time of acute disease proves to be an accurate reflection of cord blood antibody for that infant, this would make a major difference to the design of the GBS3 trial as it would negate the need for the full 180,000 cord sera samples. This would mean that we would need blood from only 150 cases of iGBS serotype III disease rather than collecting cord blood from 180,000 women. We can resolve this during the extended feasibility study.

Contributions of authors

Dr Clara Carreras-Abad (https://orcid.org/0000-0002-9068-3233) was responsible for the day-to-day operation of the study, laboratory work, data analysis and drafting the final report.

Ms Madeleine Cochet (https://orcid.org/0000-0002-0436-4843) was responsible for the study management.

Mr Tom Hall (https://orcid.org/0000-0002-9001-6684) and Ms Laxmee Ramkhelawon (https://orcid.org/0000-0001-8038-8313) were responsible for the immunological analysis of serum samples at St George’s, University of London.

Dr Asma Khalil (https://orcid.org/0000-0003-2802-7670), Dr Elisabeth Peregrine (https://orcid.org/0000-0003-4504-2979), Dr Latha Vinayakarao (https://orcid.org/0000-0003-1830-7812), Dr Sharmila Sivarajan (https://orcid.org/0000-0001-8171-6358) and Dr Rosol Hamid (https://orcid.org/0000-0001-5716-9753) (co-applicants) were principal investigators at St George’s University Hospitals NHS Foundation Trust, Kingston Hospital NHS Foundation Trust, Poole Hospital NHS Foundation Trust, Surrey and Sussex Healthcare NHS Trust, and Croydon Health Services NHS Trust, respectively.

Dr Tim Planche (https://orcid.org/0000-0002-0263-0888) contributed to the microbiologic study design and was responsible for all samples analysed at St George’s University Hospitals NHS Foundation Trust.

Dr Elizabeth Sheridan (https://orcid.org/0000-0002-7804-3217) and Dr Stephen Winchester (https://orcid.org/0000-0003-3377-3256) were in charge of all samples analysed at Poole Hospital NHS Foundation Trust and Surrey and Sussex Healthcare NHS Trust, respectively.

Mrs Jane Plumb (https://orcid.org/0000-0002-0738-3695) gave invaluable support to the study design and running.

Dr Abdelmajid Djennad (https://orcid.org/0000-0002-4830-0331) and Dr Nick Andrews (https://orcid.org/0000-0003-2069-2684) were responsible for the statistical analysis.

Dr Kirsty Le Doare (https://orcid.org/0000-0002-5104-085X) and Professor Paul Heath (https://orcid.org/0000-0002-7540-7433) had overall responsibility for the study and were involved in all stages of the work: design of the study, data analysis and drafting and commenting on the final report.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review and agreement of the Steering Committee.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care.