The yield of procalcitonin and Interleukin-6 in predicting intraamniotic infection in the presence of intrapartum fever: A pilot study

Intrapartum fever (IF) accompanied by either maternal or foetal tachycardia, elevated WBC, or purulent discharge is classified as "suspected triple 1", the hallmark of intraamniotic infection (IAI). Poor specificity of the clinical diagnosis of IAI results, in retrospect, in the unnecessary treatment of most parturients and neonates. We studied the yield of specific acute phase reactants (APRs): procalcitonin, CRP, IL-6, in detecting bacterial IAI among parturients classified as "suspected triple 1" (cases) compared to afebrile parturients (controls). Procalcitonin, CRP, and IL-6 were all significantly elevated in the cases compared to the controls, yet this by itself was not sufficient for an additive effect in detecting a bacterial infection among parturients clinically diagnosed with "suspected triple 1", as demonstrated by the poor area under the receiver operating characteristic curve of all three APRs.


Introduction
Intrapartum fever (IF), an alarming sign, is the hallmark of intraamniotic infection (IAI). Complications of IAI include maternal and neonatal sepsis, increased rates of operative vaginal deliveries and cesarean deliveries (CD), and may even result in stillbirth [1,2]. Due to its severe implications, treatment with broad spectrum antibiotics is initiated following a clinical diagnosis based on the presence of IF accompanied by either maternal or fetal tachycardia, elevated WBC, or purulent discharge, classified as "suspected triple 1"/clinical chorioamnionitis [3,4]. Poor specificity of the clinical diagnosis results, in retrospect, in the unnecessary treatment of most parturients and neonates. In the last two decades, procalcitonin (PCT) has emerged as a potential biomarker for detecting the presence of severe bacterial infections [5]. Therefore, we aimed to study the yield of PCT together with other acute phase reactants (APRs), CRP and Interleukin-6 (IL-6), in detecting bacterial IAI among parturients classified as "suspected triple 1".

Materials and methods
A prospective observational cohort pilot study was conducted at a single university hospital. Participants were recruited between June 8 th , 2020  IF classified as "suspected triple 1" (cases) and afebrile parturients in active labor (controls) were recruited to the study. All study participants were at term (37 weeks gestation and onwards). Excluded were GBS carriers, controls who developed fever after enrolment, and parturients who underwent a CD. This study was designed before the outbreak of the COVID-19 pandemic and recruitment occurred during the pandemic. However, all participants tested negative for COVID-19, in accordance with the Israeli Ministry of Health testing guidelines in place, at the time of enrolment. Clinical and laboratory implications of maternal COVID-19 infection and their effect on neonatal morbidity and mortality were previously described [6,7]. At admission, all parturients underwent a full medical history and physical examination along with a complete blood count. APRs were sampled from cases and controls at recruitment and prior to discharge from the hospital. Urine and blood cultures were obtained from all cases prior to initiation of antibiotics and from all controls at the time of recruitment. Placentas from all participants were cultured and sent for pathological assessment. We aimed to recruit 20-30 participants in each of the study groups, as is acceptable in pilot studies with unknown variance [8].
The study protocol was reviewed and approved by the local IRB (0278-19-SZMC) and all study participants provided written informed consent. The study is in accordance with the Standard for Reporting Diagnostic Accuracy (STARD), as required by the guidelines for Enhancing the QUAlity and Transparency Of health Research.

Statistical analysis
Obstetric characteristics, levels of APRs, results of maternal cultures and placental pathology comparing between cases and controls are presented as proportions, median or mean for categorical, ordinal, or continuous variables, respectively. Statistical significance was defined by a two-sided p value � 0.05 using the Chi-square test, Fisher Exact test, and Mann-Whitney U test, according to variable characteristics. A receiver operating characteristic (ROC) curve was performed to assess the yield of APRs in the diagnosis of IAI defined by (1) positive maternal bacterial cultures (urine, blood, or placenta) (2) placental histopathology confirming inflammation.
Among cases, applying a ROC analysis revealed a low, non-significant AUC for all three APRs for detecting a bacterial infection CRP 0.456, PCT 0.588 and IL-6 0.471. The detection of placental inflammation was better for IL-6 compared to CRP and PCT, AUC 0.813, 0.125 and 0.479 respectively, yet did not reach statistical significance Fig 1b and 1c.

Discussion
Among parturients with "suspected triple 1", we found a limited additive effect of APRs in the early detection of a bacterial infection, defined by either positive blood, urine, or placenta culture or the prediction of the histological report of placental inflammation. A systematic review, conducted a decade ago, identified 26 studies that evaluated the yield of serum biomarker to detect infections in "suspected triple 1" cases. PCT was studied in only one and was sampled from the newborn only [9]. A recent systematic review identified 12 studies, mostly conducted antepartum [11/12], PCT was evaluated in only 4 studies. None of these studies focused on parturients with IF [10]. CRP and IL-6 were significantly elevated in cases compared to controls and the authors concluded that these are the two biomarkers useful for detecting infection during pregnancy. In our study PCT, CRP, and IL-6 were all significantly elevated in cases compared to controls (Fig 1a), yet this by itself was not sufficient for an additive effect in detecting a bacterial infection among parturients clinically diagnosed with "suspected triple 1", as seen by the poor AUC of all three APRs. Only 4 women (15%) with "suspected triple 1" had a positive bacterial culture, therefore, 6 of seven women (~85%) will be treated with unnecessary broad-spectrum antibiotics and their neonates will be exposed to invasive procedures and unnecessary antibiotics.
This study has a few strengths. While most studies have examined the yield of PCT in the detection of IAI in the preterm period and in cases of preterm premature rupture of membranes [11][12][13], this study included only term pregnancies during labor. There are many known factors that increase the risk of preterm delivery, such as maternal age [14] and infectious conditions [15], potentially confounding the results of a study examining the yield of APRs. Additionally. This study was a controlled comparative study, lowering the risk of bias.
Limitations of this study include the small number of participants, which may have been too small to detect the value of APRs in the diagnosis of IAI. There is no gold standard for the diagnosis of IAI and thus there is no way to ascertain that all of the IAI group participants did, in fact, have the diagnosis. Future studied are needed to ascertain factors that could better identify bacterial induced IF.