Abstract
Preterm birth is a major cause of perinatal mortality and morbidity in newborns, and its risk is increased by bacterial vaginosis (BV) during pregnancy. This multicenter prospective cohort study aimed to evaluate whether Flora select™ (FS), a newly developed real-time polymerase chain reaction test, is clinically useful for assessing the effectiveness of BV treatment during pregnancy. The vaginal microbiome characterized by relative dominance rates of Lactobacillus ≤ low (<50%), together with a positive test for Gardnerella, Prevotella, or Atopobium species, was defined as BV-FS A criterion. The vaginal microbiome characterized by Lactobacillus medium (50%≤, <80%), together with positive tests for Gardnerella plus either Prevotella species or Atopobium species, was defined as BV-FS B criterion. This study enrolled 25 pregnant women with classical BV (Nugent score ≥ 7) at initial examinations, and they met the BV-FS A (n = 23) and BV-FS B (n = 2) criteria. No woman with classical BV had a missed diagnosis of molecular BV. Treatments with metronidazole vaginal tablets resulted in the improvement of 88.0% (22/25) of classical BV, 65.2% (15/23) of BV-FS A, and 50.0% (1/2) of BV-FS B cases, whereas positive rates of Ureaplasma species in women with classical BV increased by 42.9%. Although most classical BV cases were cured following metronidazole treatments, a considerable proportion still harbored molecular BV detected by FS. Although the Nugent scoring system revealed that 80.0% (20/25) of women with classical BV (Nugent score ≥ 7) were sufficiently cured as BV-negative (Nugent scores 0-3), 5 (25%) of the 20 cured cases still met the BV-FS A/B criteria. FS particularly detected Ureaplasma species in 9 (45%) of the 20 cured cases. It could identify pregnant women who require additional treatments for residual molecular BV and Ureaplasma species. Therefore, the FS test may be clinically useful for assessing the vaginal microbiome and evaluating the effectiveness of BV treatments.