Microbiology substudy

Parts of this chapter have been reproduced from Daniels et al.1 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

Substudy objectives

• To determine the antibiotic resistance profile of any GBS isolated from the rectal vaginal swab taken from the mother at the time of labour, and to compare this with the antibiotic resistance profile from any faecal sample taken from the woman’s baby at 6 weeks of age.
• To estimate the carriage rate of antibiotic-resistant bacteria, particularly E. coli, in rectal samples from women recruited to the GBS2 trial from centres in LSE who are assigned a rapid test system. In addition to antibiotic-resistant E. coli, we have also included MRSA and VRE.
• To estimate the extent to which carriage of specific resistant bacteria or resistance elements by the mother at the time of birth increases the risk of carriage of those specific bacteria or elements by the infant at 6 weeks of postnatal age.
• To gather some information on peripartum risk factors for transmission (e.g. mode of birth, maternal comorbidities, gestational age and antibiotic exposure).

Methods

The substudy recruited only in the participating centres in LSE that were randomised to receive a rapid test system. Written consent was sought for the 6-week follow-up of the children born to study participants in the participating centres. Vaginal/rectal swabs were taken from women who had risk factors associated with early-onset GBS infection around the time of birth using a third swab. This swab was taken alongside or directly after the first and second swabs, as in other centres assigned a rapid test (see Chapter 7).

Additional microbiological assessment of the third vaginal/rectal swab

The third swab was sent to the microbiology laboratory at Barts Health NHS Trust and stored in a Brain Heart Infusion Broth (Oxoid Ltd, Basingstoke, UK) with 10% glycerol cryopreservative broth at –80 °C. This swab was processed further only if notification of written consent for substudy participation had been obtained from the mother. The third swab was held for a period of 96 hours in the microbiology department of Barts Health NHS Trust to allow the woman to receive an invitation to participate in research and give informed consent. Should a woman have declined to provide consent, or if 96 hours had passed since the receipt of the third swab, then this third swab was not processed but instead was disposed of in a suitable manner compliant with local policies.

Obtaining a faecal sample from the infant

Once a period of 5 weeks had elapsed and following notification from the trial office that consent had been given, the substudy research assistant supplied the maternity unit with a faecal sample pot labelled with the baby’s unique study number. The research assistant then prompted the dispatch of a follow-up sample collection pack to the consenting mother. The research assistant recorded the number on the pot alongside the mother’s trial number. The maternity unit staff or research midwife/practitioner then checked with the community midwifery team that nothing untoward had happened to the child, then the research midwife/practitioner sent out a follow-up sample collection pack. This pack consisted of a faecal sample pot, a covering letter, instructions, a protective transport container and a prepaid addressed envelope to the woman’s home address. The letter requested that a sample of the child’s faecal material be collected from the child’s nappy. The instructions provided to the mother are shown in Report Supplementary Material 1.

When a sample of the baby’s faecal material was not forthcoming after the first request, then another request was sent to the mother at around 9 weeks after her baby was born. No further attempts to obtain a faecal sample were made.

Microbiological testing of infant faecal samples

Methods for the isolation of GBS are described elsewhere in this monograph. Following receipt of the infant faecal sample, the microbiology laboratory at Barts Health NHS trust informed the research assistant, who then recorded this against the woman’s unique trial number. The faecal sample was diluted 1 in 10 into a cryopreservation broth and then mixed by vortexing, before being stored at –80 °C.

Antibiotic resistance testing

The frozen sample was allowed to defrost at room temperature and then plated on to a range of culture media, including chromogenic MRSA agar (Oxoid Ltd), Slanetz–Bartley agar with vancomycin (Oxoid Ltd) and MacConkey agar (Oxoid Ltd). Presumptive isolates were identified using matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry. Antibiotic resistance was tested using European Committee on Antimicrobial Susceptibility Testing methods and break points.73 Enterobacterales were tested for sensitivity to ampicillin, piperacillin plus tazobactam, amoxicillin plus clavulanate, cefpodoxime, gentamicin, cefuroxime, amikacin, co-trimoxazole, temocillin, ceftazidime, ertapenem and ciprofloxacin on Muller–Hinton agar.

Molecular testing for Gram-negative antibiotic resistance genes

The EasyScreen™ Sample Processing Kit (SP006, Genetic Signatures Ltd, Newtown, NSW, Australia) is designed to rapidly isolate nucleic acids [deoxyribonucleic acid (DNA) and ribonucleic acid (RNA)] from clinical samples via an automated purification system. The system was used to look for antibiotic-resistant genes prevalent in Gram-negative bacteria. Reagents are supplied to convert all nucleic acids into 3base™ sequences (Genetic Signatures Ltd) and to perform a subsequent purification on the GS-mini platform (Genetic Signatures Ltd). The original maternal vaginal swab was removed from the –80 °C freezer and vortexed briefly in the Amies transport medium in which it was stored to release the bacteria. A 1-ml aliquot was removed and added to the GSL-combined reagents 1 and 2. Infant samples were removed from the –80 °C freezer and, after defrosting, 100 µl/ml was added to combined GSL reagents 1 and 2 for processing with the EasyScreen™ Sample Processing Kit (SP006).

The EasyScreen™ Sample Processing Kit lyses any microorganisms present and converts all cytosine bases to uracil (detected as thymine after PCR amplification) to create 3base™ DNA and RNA. The 3base™ nucleic acids have an increased homology compared with the genome that comprised the native four bases. This complexity reduction results in genomes that are more similar to each other, enabling the design of primers and probes that contain fewer mismatches, are more homologous, produce better amplification and, importantly, have less cross-reactivity when multiple strains of the same organism are present.

Extracted DNA was analysed using the EasyScreen™ ESBL/CPO Detection Kit (BL001, Genetic Signatures Ltd), which uses panels of multiplexed real-time PCR assays to provide rapid and accurate detection of 16 β-lactam and carbapenem-resistant pathogen targets. These were TEM, GES, OXA-48(like), MCR-1, VIM, CMY, CTX-M, KPC, OXA-23 (like), DHA, IMI, IMP, NDM, SHV, SME and OXA-51 (like). The kit includes an extraction control in the reaction mix to determine the reliability of the extracted nucleic acids and indicate the presence of any inhibitors after extraction from primary samples. PCR was performed on a Bio-Rad CFX96/384 Touch™ real-time PCR instrument. Reaction mixtures were set up in accordance with the manufacturer’s instructions and C_t values read in accordance with its recommendations.

Statistical methods

A binomial regression model with a log-link was used to estimate RRs. All estimates are reported along with 95% CIs. All analyses were conducted in SAS^® version 9.4 (SAS Institute Inc., Cary, NC, USA). Fisher’s exact test was used to compare proportions and a p-value of < 0.05 was considered statistically significant.

Results

Overall, 117 women provided samples for the substudy. There were 64 paired samples from 63 mothers and 64 infants (one set of twins). There were 60 paired samples available for molecular testing (59 mothers and 60 infants).

Group B streptococci were isolated from 39 (33%) of the 117 maternal samples. Of these, 32 out of 39 (82%, 95% CI 66% to 92%) isolates were tetracycline resistant, 9 out of 39 (23.1%, 95% CI 11% to 39%) were erythromycin resistant and 7 out of 39 (18%, 95% CI 8% to 34%) were clindamycin resistant, but there were no penicillin-resistant GBS isolates (0%, 95% CI 0% to 9%). GBS were isolated from only 2 of the 64 infant 6-week samples (3%, 95% CI 0% to 11%). There were no MRSA isolated from mother or infant samples. VRE were isolated from one infant and one unrelated maternal sample.

E. coli were isolated from 85 (73%) of the 117 maternal samples. Of these, resistance to ampicillin was found in 46 (54%) of the 85 samples, resistance to amoxycillin/clavulanate was found in 37 (44%) samples, resistance to trimethoprim/sulfamethozaxole was found in 21 (25%) samples, resistance to ciprofloxacin was found in five (6%) samples and resistance to gentamicin was found in four (5%) samples. ESBL-producing bacteria were cultured from three (4%) samples, and 18 (21%) samples were resistant to three or more antibiotic classes (multiresistant).

Gram-negative antibiotic resistance genes were identified in the 128 samples from mother–infant pairs. Genes identified in samples included TEM, CTX-M, SHV, OXA-23, 48, 51 (like), CMY, IMI, VIM, MCR-1, DHA, GES, KPC and NDM. Only TEM and CTX-M were present in ≥ 10 maternal samples. SHV (which is carried by Klebsiella pneumoniae predominantly) was found in 14 out of 60 (23%) infant samples compared with only 5 out of 59 (8%) maternal samples, possibly reflecting differences in mother and infant Klebsiella colonisation (Fisher’s exact test p = 0.04).

The RR of carriage of resistant E. coli or resistance genes in infants born to mothers with or without carriage of strains with specific characteristics is shown in Table 30. Although the point estimates for association between infants carrying any resistant E. coli or specific resistance genes were consistently higher when carriage was detected in the mother, the findings were significant only for co-trimoxazole-resistant and multidrug-resistant genes.

In the adjusted analysis that included mode of birth, gestational age, neonatal and maternal peripartum antibiotic exposure, and neonatal intensive care admission, small numbers precluded the convergence of a ‘full’ model (i.e. one adjusted for all five variables) and, therefore, there are no estimates of risk ratios or CIs taking account of all variables. Instead, all adjusted analyses included a single adjustment variable. The results of the analyses are shown in Table 31, in which the analyses in each column are adjusted by only the single adjustment variable indicated in the column heading.