First-Trimester Serum Cytokine Profile in Pregnancies Conceived After Assisted Reproductive Technology (ART) With Subsequent Pregnancy-Induced Hypertension

Xiangxin Lan, Ling Guo, Shiqin Zhu, Yongzhi Cao, Yue Niu, Shuwen Han, Zeyan Li, Yan Li, Junhao YanFrontiers in Immunology 2022 · DOI

TL;DR

Researchers have identified a distinct "downregulated" cytokine signature in the first-trimester serum of women who conceive via Assisted Reproductive Technology (ART) and later develop pregnancy-induced hypertension (PIH). While classical biomarkers like PlGF and sFLT1 fail to predict PIH early in ART pregnancies, a combined model of three specific cytokines—IL-7, MIG, and SCF—offers a promising new tool for early risk assessment.

What Problem Does This Paper Address?

Pregnancy-induced hypertension (PIH) is a major complication affecting 6–10% of pregnancies, characterized by high blood pressure and potential multi-organ damage after the 20th week of gestation. Crucially, women who conceive via ART (like IVF or ICSI) face a significantly higher risk of PIH than those with spontaneous pregnancies.

The current clinical challenge is that "classical" predictive markers—such as placental growth factor (PlGF) and soluble Fms-like tyrosine kinase-1 (sFlt-1)—have poor predictive accuracy during the first trimester, especially in the ART population. Because early intervention (like low-dose aspirin) is most effective when started before 14–16 weeks, there is an urgent need for biomarkers that can identify high-risk ART pregnancies much earlier in the first trimester.

Methods

The researchers conducted a case-control study using serum samples collected at 11–13 gestational weeks from 66 women who conceived via ART: 33 who later developed PIH and 33 matched normotensive controls (NC).

Key methodological steps included:

  • Propensity Score Matching (PSM): Used to balance baseline characteristics like age, BMI, and cause of infertility between the two groups.
  • Multiplex Immunoassay: They used a Bio-Plex Pro Human Cytokine 48-plex panel to measure a wide array of cytokines, chemokines, and growth factors, alongside standard ELISA for classical markers (Activin A, PlGF, sFLT1).
  • Bioinformatic Pipeline: Analysis involved Principal Component Analysis (PCA) to see if groups could be distinguished by their profiles, heatmap/cluster analysis, and univariate logistic regression to find correlations with PIH onset.
  • Model Construction: They used Receiver Operating Characteristic (ROC) analysis to determine the predictive power of individual cytokines and built a multi-factor combined model.

Datasets

The study utilized a primary clinical dataset of 66 participants recruited from a large prospective cohort at the Research Center of the Reproductive Hospital Affiliated to Shandong University, China, between January 2018 and June 2019. No external secondary datasets were used for training or validation.

Key Insights and Ideas

  • The "Downregulation" Paradox: While PIH in later stages is typically associated with a pro-inflammatory state (high cytokines), this study found that in the first trimester, 29 different cytokines were significantly lower in the PIH group. This suggests that PIH in ART might stem from an initial failure of immune activation or "tolerance induction" rather than early-stage inflammation.
  • Failure of the "Gold Standards": The study confirmed that classical markers (Activin A, PlGF, sFLT1) showed no significant difference between groups at 11–13 weeks, highlighting their inadequacy for early screening in ART populations.
  • The ART-Immune Link: The authors hypothesize that ART pregnancies may lack the "immune priming" that occurs during natural conception (e.g., exposure to seminal fluid), leading to the observed lower cytokine levels and subsequent poor placental development.
  • Predictive Power of the "Big Three": Out of 17 cytokines negatively correlated with PIH, three stood out: IL-7, MIG, and SCF. Combining these into a single model significantly outperformed any individual marker.

Key Results

  • Cytokine Profile: 29 cytokines were significantly lower in the PIH group ($P < 0.05$).
  • Correlation: 17 cytokines (including IL-2, IL-6, IL-10, and G-CSF) were significantly negatively correlated with the subsequent onset of PIH.
  • Model Performance: The combined model (IL-7 + MIG + SCF) achieved an Area Under the Curve (AUC) of 0.821 (95% CI: 0.718–0.924, $P < 0.001$).
  • Sensitivity/Specificity: The model demonstrated a sensitivity of 81.8% and a specificity of 72.7% for predicting PIH.

Limitations and Future Work

The authors acknowledge several limitations:

  • Sample Size: The study was limited to 33 pairs (66 total participants), which is relatively small for a biomarker discovery study.
  • Single-Center Study: All participants were from one hospital in China, meaning the results may not generalize to other ethnic or geographic populations.
  • Lack of Mechanistic Depth: While the study identifies a correlation, it does not prove that low cytokine levels cause PIH.

Future Work: The researchers suggest that larger, multi-center prospective studies are needed to validate this cytokine model. Additionally, further research into the "immune tolerance" hypothesis could explain why ART-conceived embryos face different immunological hurdles than those conceived spontaneously.

Summary updated

Abstract

Pregnancy-induced hypertension (PIH) is one of the most common pregnancy complications that seriously affects the mother and fetus. The incidence of PIH is higher in pregnancies conceived after assisted reproductive technology (ART) than in spontaneous pregnancies; thus, exploring potential serum biomarkers before PIH onset is of great significance for effective early prediction and prevention of PIH in the ART population. Cytokines are involved in the inflammatory response and immune regulation, which play an essential role in the pathogenesis of PIH. A description of the cytokine profile in the first trimester of pregnancy could help identify new diagnostic tools and develop targeted therapies for PIH in the ART population. The concentrations of classical predictive markers for PIH and another 48 cytokines were measured in the first-trimester pregnancy serum samples from 33 PIH patients and 33 matched normotensive controls (NC), both of whom conceived after ART treatment. The measured values were compared and analyzed between NC and PIH, followed by comprehensive bioinformatic analysis and logistic regression analysis. There was no significant difference in classical predictive markers, including Activin A, PlGF, sFLT1 (VEGFR), and sFLT1/PlGF, between the PIH and NC groups (<i>P</i> > 0.05), while 29 cytokines were significantly lower in the PIH group than in the NC group (<i>P</i> < 0.05). Logistic regression analysis revealed that 17 cytokines (IL-2Rα, M-CSF, IL-6, IL-2, β-NGF, IL-7, IL-12 (p70), SCF, IL-10, IL-9, MIG, GM-CSF, LIF, IL-1α, MCP-3, IL-4, and HGF) in the first-trimester pregnancy serum were significantly negatively correlated with the subsequent onset of PIH. With the top 3 cytokines (IL-7, MIG, and SCF) of receiver operating characteristic (ROC) analysis, we constructed an efficient multifactor combined detection and prediction model for PIH in ART pregnancy. Classical early predictors for hypertensive disorder complicating pregnancy cannot distinguish PIH from their normal peers in ART pregnancy. In comparison, the description of the cytokine profile in the first trimester of pregnancy enables us to distinguish high-risk ART pregnancy for PIH, permitting enough time for PIH prevention therapy. The cytokine profile we described also provides immunological insight into the further mechanistic exploration of PIH.
BibTeX
@inproceedings{lan2022first,
  author = {Xiangxin Lan and Ling Guo and Shiqin Zhu and Yongzhi Cao and Yue Niu and Shuwen Han and Zeyan Li and Yan Li and Junhao Yan},
  title = {{First-Trimester Serum Cytokine Profile in Pregnancies Conceived After Assisted Reproductive Technology (ART) With Subsequent Pregnancy-Induced Hypertension}},
  booktitle = {Frontiers in Immunology},
  year = {2022},
  doi = {10.3389/fimmu.2022.930582},
}
Figure