080 - Bacterial DNA load of amniotic membrane by digital droplet PCR is associated with histologic chorioamnionitis

Publication Number: 80.3748

Nikou P.. Chokshi, Endeavor Health, Chicago, IL, United States; Aspen N. Kremer, University of Chicago Division of the Biological Sciences The Pritzker School of Medicine, Chicago, IL, United States; Delia N. Mwinyelle, Universiry of Chicago, Indianapolis, IN, United States; Jeffery Goldstein, Northwestern University The Feinberg School of Medicine, Chicago, IL, United States; Sebastian Otero, Comer Children's Hospital at University of Chicago Medical Center, Chicago, IL, United States; Patrick C. Seed, Northwestern University, Evanston, IL, United States; Leena B. Mithal, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States

Background: Chorioamnionitis (chorio) refers to intraamniotic inflammation and increases risk of neonatal infection. Clinical chorio aligns moderately with placental histopathology. Availability of placental histopathology results and microbiologic cultures are often beyond the timeframe to impact clinical care. Rapid molecular diagnostics may optimize antibiotic decision-making in the setting of clinical chorio. We hypothesized that bacterial DNA load, measured by digital droplet PCR (ddPCR), would be elevated in placentas affected by chorioamnionitis.

Objective: To determine the bacterial 16S rDNA load of amniotic membrane (amnion) and chorionic plate (chorion) tissue by ddPCR among pregnancies with and without 1) clinical and 2) histologic chorioamnionitis.

Design/Methods: Fifty-two placentas from full-term deliveries were included in this study (n=25 clinical chorio, n=27 controls). The study was approved (exempt) by Northwestern University IRB. Placental samples were formalin fixed, paraffin embedded, and underwent histopathologic analysis with grading of maternal inflammatory response (MIR; Amsterdam criteria, 0-3 scale). Histologic MIR was coded and analyzed as continuous and dichotomous (low 0-1, high 1.5-3) variables. Separate biopsies of chorion and amnion were obtained, weighed, and stored at -80°C. DNA was isolated from 70-80mg of tissue. A universal Taqman probe targeting the 16S rRNA gene was used on the QX200 ddPCR to quantify bacterial load for each sample in triplicate. Correlation and receiver operating characteristics analyses were conducted in STATA.

Results: Clinical chorio was diagnosed by criteria in Table 1; 71% of cases had maternal fever alone. Higher bacterial DNA load in amnion was significantly associated with MIR of amniotic membrane (R=0.39, p=0.005) and nearly reached significance with MIR of chorion (R=0.2734, p=0.052) but did not correlate with clinical chorio [Table 2]. Bacterial load of chorion was not associated with MIR of placental tissues or clinical chorio. Bacterial load of amnion as a predictor of low or high MIR of amnion had an AUC of 0.72 [Figure 1]. Using Youden’s method of establishing a cut point, a threshold of 7.7copies/uL to diagnose high MIR of amniotic membrane yielded an AUC of 0.69.

Conclusion(s): There is a significant association between increasing amnion bacterial load detected by ddPCR in placentas with histologic chorio/MIR. These findings are plausible given pathophysiology of an ascending infection as an inflammatory trigger and may be clinically useful as a rapid, ancillary placental diagnostic tool. Further validation across larger cohorts is necessary.

Demographic and Clinical Characteristics


Association of bacterial DNA load by ddPCR with histologic and clinical chorioamnionitis
MIR = maternal inflammatory response on placental histopathology as continuous variable (0-3)

ROC curve for bacterial DNA load in amniotic membrane tissue and maternal inflammatory response
Bacterial DNA load by ddPCR of amnion biopsy sample as predictor of histopathologic maternal inflammatory response (MIR, low 0-1 vs. high 1.5-3)