RESUMO
OBJECTIVE: This study was aimed to develop models using multiple cytokine/chemokine levels in cervicovaginal fluid (CVF) and plasma and widely used noninvasive parameters that have better accuracy for predicting intra-amniotic infection and/or inflammation (IAI) and imminent spontaneous preterm delivery (SPTD, ≤48 hours) in women with preterm labor (PTL). STUDY DESIGN: This was a retrospective cohort study of 95 singleton pregnant women with PTL (23-34 weeks) who underwent amniocentesis. Both CVF and plasma samples were obtained at the time of amniocentesis, and serum C-reactive protein (CRP) levels were measured. The amniotic fluid (AF), CVF, and plasma samples were assayed for interleukin (IL)-6, IL-8, IL-10, monocyte chemotactic protein-1 (MCP-1), and macrophage inflammatory protein-1ß (MIP-1ß) levels using a multiplex immunoassay kit. RESULTS: The levels of most cytokines/chemokines measured in the AF and CVF were significantly higher in the women with than in those without IAI and imminent SPTD, whereas only high-plasma IL-10 level showed a significant association with imminent SPTD. In predicting IAI, proteins in AF had significantly higher areas under the curves (AUCs) than those in CVF and plasma. However, for predicting imminent SPTD, no significant differences in the AUCs of the outcome-associated proteins were observed among the measurements in AF, CVF, and maternal plasma. By using stepwise regression analyses, noninvasive models (using protein levels in CVF and baseline clinical parameters) were developed for the prediction of IAI and imminent SPTD. The AUC of these noninvasive models were similar to those of the invasive models (using AF protein levels and baseline clinical parameters). CONCLUSION: Noninvasive models based on CVF cytokine/chemokine levels and widely used noninvasive parameters (especially CRP) act as good indicators for predicting the risk of IAI and imminent SPTD in women with PTL. Evaluation of cytokine/chemokine levels in plasma samples did not add valuable information regarding the two outcome measures in the PTL setting. KEY POINTS: · Markers in either CVF or plasma alone did not have sufficient accuracy for predicting IAI and SPTD.. · Noninvasive models using CVF cytokine and CRP act as effective tools for predicting two outcomes.. · Evaluation of cytokine level in plasma did not add valuable information regarding two outcomes..
