Project Details

Early Career

Status: Funded - Closed

Cord blood biomarkers and 16S rRNA PCR for early diagnosis of Sepsis in premature infants

Leena Mithal, M.D.


BACKGROUND:  Despite advances in maternal screening and antibiotic prophylaxis, early onset bacterial sepsis (EOS) remains a notable cause of infant morbidity and mortality worldwide. EOS is disproportionately represented in preterm infants (<37 weeks gestation), with 47% of EOS cases and 92% of EOS deaths occurring in this group.

GAP:  Current means of EOS diagnosis are inadequate, leading to serious adverse consequences from missed cases of sepsis and prolonged empiric antibiotic treatment. The utility of umbilical cord blood immune biomarkers, PCR detection of bacterial genomes, and placental histopathology for timely diagnosis of EOS is unknown and requires further investigation.

HYPOTHESIS:  We hypothesize that certain immune biomarkers (procalcitonin [PCT], C-reactive protein [CRP], haptoglobin [Hp], ferritin, fibrinogen, tissue plasminogen activator [TPA], serum amyloid A [SAA], serum amyloid P [SAP], and α-2-macroglobulin) are elevated in cord blood of infants with EOS and that cord blood 16S rRNA PCR is predictive of early onset bacteremia. Evaluation of biomarkers in combination with placental histopathologic acute inflammation (AI) may lead to improved EOS detection. We hypothesize that correlation between immune biomarkers and positive PCR will enhance the predictive power for identification of EOS.

METHODS:  This is a nested case-control study of premature infants enrolled in a longitudinal birth cohort with archived cord blood collected by sterile venipuncture at delivery. Infants were categorized into designated sepsis categories based on culture, laboratory, clinical, and antibiotic treatment data (confirmed EOS [cEOS] n=12, “presumed” EOS [PS] n=30, late onset sepsis [LOS] n=15, and negative controls n=30). We examined nine acute phase reactant proteins using multiplex immunoassays (Bio-Rad) and bacterial PCR of cord blood of infants taking into account placental inflammation and other important maternal and infant covariates. After DNA extraction, 16S rRNA PCR, and Sanger sequencing, the NCBI database was used for species identification.

RESULTS: Causative organisms of EOS were E coli, GBS, Proteus, H influenzae and Listeria. Mean gestational age of the study patients was GA 29.7 (SD± 2.8) weeks and median birth weight 1245 g (IQR: 995-1670). Gender, multiple gestation, preeclampsia and prolonged rupture of membranes were not significantly different across groups. Cord blood CRP, SAA, Hp, and ferritin were significantly elevated in the cEOS compared to control, PS, and LOS groups (p<0.01). PCT, fibrinogen, α-2-macroglobin and TPA were not different across groups. SAA, CRP, and Hp demonstrated AUCs of 99%, 96%, and 95% respectively. Placental fetal AI was present in all cEOS patients and is associated with sepsis. Regression analysis showed significant beta coefficients for SAA, CRP, and Hp with EOS, including after adjustment for covariates including fetal AI (p<0.03). 
16S rRNA PCR results were that all 12 cEOS had a positive band on electrophoresis gel; 9 had single species identified by Sanger sequencing; 8 of these were the same organism as in postnatal blood culture. Bacterial DNA was detected in 8/30 PS (GBS, E coli, and others) and 7/30 controls (Lactobacillus, Mycoplasma, Prevotella). PCR results correlated with elevated SAA, CRP, and Hp in cEOS and PS groups. Deep sequencing of PCR amplicons detected EOS pathogens and other bacteria associated with vaginal, oral and gastrointestinal flora in sepsis and control groups.

CONCLUSIONS: Certain acute phase reactants (SAA, CRP, and Hp) are elevated in cord blood of preterm infants with EOS of intrauterine origin. Cord blood biomarker levels with consideration of placental acute inflammation may improve the detection and management of EOS. Cord blood 16S rRNA PCR identifies EOS pathogens, including in some culture-negative cases. PCR correlates with abnormal biomarkers and together may distinguish PS cases without true infection. Further investigation is needed to further elucidate the interplay between placenta, cord blood biomarkers, and EOS.  According to our preliminary deep sequencing data, cord blood may not be “sterile”. The significance of bacterial DNA in neonatal disease is unclear. Further investigation of molecular diagnostics for EOS and possible maternal-fetal origins of the infant microbiome is warranted.

IMPACT:  There is a potential role of cord blood biomarkers and 16S rRNA PCR as diagnostic tools for EOS and to help guide antimicrobial therapy in preterm infants.  Testing of cord blood could decrease mortality from severe sepsis, allow targeted antibacterial treatment, and improve overuse of empiric antibiotics leading to the development of antibiotic resistance and other adverse outcomes.

Supervising Institution:
Lurie Children's Hospital of Chicago

Ram Yogev

Project Location:

Award Amount: