Limited relationship between cervico-vaginal fluid cytokine profiles and cervical shortening in women at high risk of spontaneous preterm birth

Abstract

Objective: To determine the relationship between high vaginal pro-inflammatory cytokines and cervical shortening in women at high risk of spontaneous preterm labor and to assess the influence of cervical cerclage and vaginal progesterone on this relationship.

Methods: This prospective longitudinal observational study assessed 112 women with at least one previous preterm delivery between 16 and 34 weeks' gestation. Transvaginal cervical length was measured and cervico-vaginal fluid sampled every two weeks until 28 weeks. If the cervix shortened (<25 mm) before 24 weeks' gestation, women (cases) were randomly assigned to cerclage or progesterone and sampled weekly. Cytokine concentrations were measured in a subset of cervico-vaginal fluid samples (n = 477 from 78 women) by 11-plex fluid-phase immunoassay.

Results: All 11 inflammatory cytokines investigated were detected in cervico-vaginal fluid from women at high risk of preterm birth, irrespective of later cervical shortening. At less than 24 weeks' gestation and prior to intervention, women destined to develop a short cervix (n = 36) exhibited higher cervico-vaginal concentrations than controls (n = 42) of granulocyte-macrophage colony-stimulating factor [(GM-CSF) 16.2 fold increase, confidence interval (CI) 1.8-147; p = 0.01] and monocyte chemotactic protein-1 [(MCP-1) 4.8, CI 1.0-23.0; p = 0.05]. Other cytokines were similar between cases and controls. Progesterone treatment did not suppress cytokine concentrations. Interleukin (IL)-6, IL-8, granulocyte colony-stimulating factor (G-CSF), interferon (IFN)-γ and tumour necrosis factor (TNF)-α concentrations were higher following randomization to cerclage versus progesterone (p<0.05). Cerclage, but not progesterone treatment, was followed by a significant increase in cervical length [mean 11.4 mm, CI 5.0-17.7; p<0.001].

Conclusions: Although GM-CSF and MCP-1 cervico-vaginal fluid concentrations were raised, the majority of cervico-vaginal cytokines did not increase in association with cervical shortening. Progesterone treatment showed no significant anti-inflammation action on cytokine concentrations. Cerclage insertion was associated with an increase in the majority of inflammatory markers and cervical length.

Figure 1. Enrolment, randomization and follow-up of participants in the study.

Figure 2. Kaplan-Meier plot to compare time to delivery following allocation to treatment with progesterone (solid line, n = 17) and cervical cerclage (dashed line, n = 19); p = 0.23.

Vaginal progesterone was prescribed on the day of allocation; women with cerclage received treatment within 2.5 (SD 3.9) days.

Figure 3. Representative gestational profile of 11 cytokines measured longitudinally in cervico-vaginal fluid samples from a pregnant woman at risk of spontaneous preterm labour who did not demonstrate cervical shortening or deliver preterm.

A. Interleukin (IL)-4 (solid line, closed circle); IL-7 (dotted line, open triangle); granulocyte-macrophage colony-stimulating factor (GM-CSF) (dashed line, closed square). B. IL-1β (solid line, closed circle); interferon (IFN)-γ (dotted line, open triangle); macrophage inflammatory protein (MIP)-1β (dashed line, closed square). C. IL-8 (solid line, closed circle); IL-6 (dotted line, open triangle) circle); granulocyte colony-stimulating factor (G-CSF) (dashed line, closed square). D. Monocyte chemotactic protein (MCP)-1 (solid line, closed circle) and tumour necrosis factor (TNF)-α (dotted line, open triangle).

Figure 4. Representative gestational profile of 11 cytokines measured longitudinally in cervico-vaginal fluid samples from a representative woman at risk of spontaneous preterm labor who demonstrated cervical shortening and randomised to receive a cervical cerclage at 21⁺⁴ weeks’ gestation and delivered preterm at 32 weeks’ gestation.

A. Interleukin (IL)-4 (solid line, closed circle); IL-7 (dotted line, open triangle); granulocyte-macrophage colony-stimulating factor (GM-CSF) (dashed line, closed square). B. IL-1β (solid line, closed circle); interferon (IFN)-γ (dotted line, open triangle); macrophage inflammatory protein (MIP)-1β (dashed line, closed square). C. IL-8 (solid line, closed circle); IL-6 (dotted line, open triangle) circle); granulocyte colony-stimulating factor (G-CSF) (dashed line, closed square). D. Monocyte chemotactic protein (MCP)-1 (solid line, closed circle) and tumour necrosis factor (TNF)-α (dotted line, open triangle).

Figure 5. Gestational profiles of 11 cytokines measured longitudinally in cervico-vaginal fluid samples from a representative woman at risk of spontaneous preterm labor who demonstrated cervical shortening, was allocated to receive vaginal progesterone after the 23⁺¹ weeks’ gestation CVF sample was taken, and who delivered preterm at 30⁺¹ weeks’ gestation.

A. Interleukin (IL)-4 (solid line, closed circle); IL-7 (dotted line, open triangle); granulocyte-macrophage colony-stimulating factor (GM-CSF) (dashed line, closed square). B. IL-1β (solid line, closed circle); interferon (IFN)-γ (dotted line, open triangle); macrophage inflammatory protein (MIP)-1β (dashed line, closed square). C. IL-8 (solid line, closed circle); IL-6 (dotted line, open triangle) circle); granulocyte colony-stimulating factor (G-CSF) (dashed line, closed square). D. Monocyte chemotactic protein (MCP)-1 (solid line, closed circle) and tumour necrosis factor (TNF)-α (dotted line, open triangle).

Figure 6. Comparison of cervical length, measured by transvaginal ultrasound in vivo, in controls and in women with a short cervix who were treated with a cervical cerclage or progesterone when the cervix shortened to ≤ 25 mm.

A. Cervical length measurements versus the gestation at which measurements were obtained (controls, solid line, closed circles; cerclage group, dashed line, closed triangle and vaginal progesterone, dashed line, closed circle). B. Cervical length measurements versus the time from when cervical shortening was first detected (cerclage group, dashed line, closed triangle and vaginal progesterone, dashed line, closed circle). Cervical length measurements for controls are shown against gestational age (solid line, closed circles).