Biomarkers for predicting spontaneous preterm birth: an umbrella systematic review

Objective: To identify all systematic reviews investigating the role of maternal and fetal biomarkers for predicting spontaneous preterm birth (SPTB). Methods: Medline and Web of Sciences databases were searched electronically. Studies explor- ing the association between maternal biomarkers and spontaneous delivery were considered suitable for inclusion. A synthesis of the systematic reviews was performed with the umbrella methodology. Statistical measures of association (Odd ratio, OR, relative risk, RR) and predictive accuracy (sensitivity, specificity, positive and negative likelihood ratios were used to synthesize results of the included studies. Results: 21,614 articles were identified, 542 were assessed with respect to their eligibility for inclusion and 14 systematic reviews included. Cervical fibronectin was the biomarkers which showed the highest strength of association with the occurrence of SPTB (delivery within 24 h OR 7, 95%CI 3–17; delivery <7days (OR 12, 95%CI 8–16). Maternal serum alpha fetoprotein, was associated with an OR of 4 and 3 for early and late SPTB. C-reactive protein had an OR of 2 (95%CI 1–2) and 8 (95%CI 4–16) when detected in maternal plasma and amniotic fluid, respect- ively. Among cytokines, interleukin-6 had an OR and an LRþfor SPTB of 2 and 12 when detected in maternal serum. Conclusions: Cervical fetal fibronectin, alpha fetoprotein, C- reactive protein and interleukin 6 can have an overall good diagnostic accuracy in identifying pregnancies at risk of SPTB. Large prospective studies in different sub-set of women are needed to ascertain whether the combin- ation of different serological and imaging marker can improve antenatal prediction of this condition.