Background/Purpose: Pregnancy is known to induce a natural improvement of Rheumatoid Arthritis (RA) symptoms in 50-75% of patients as gestation progresses. However, the underlying mechanisms are not well understood and no biomarkers have been identified that predict whether a woman will improve or worsen during pregnancy. In this study, we aimed to identify RA-associated pre-pregnancy gene expression signatures to determine if they correlated with the subsequent improvement or worsening of RA during pregnancy.

Methods: Women with RA (n=11) and healthy women (n=5) were enrolled in our pregnancy cohort and followed prospectively. The women with RA were examined before pregnancy (T0) and at the third trimester (T3). Disease activity was assessed using Clinical Disease Activity Index (CDAI) scores. Blood was drawn from all women (RA and healthy) at both time-points. Total RNA was isolated and used to prepare cDNA libraries which were sequenced at an average depth of 60 million reads. The raw RNA sequencing (RNA-seq) data were pseudo-aligned to the reference Human transcriptome and quantified using kallisto. Genes differentially expressed between groups were identified with edgeR using a fold-change cutoff of 2 and significance threshold q< 0.05 (FDR corrected). Functional enrichment analysis was performed using Cytoscape.

Results: Of the 11 women with RA, 8 improved (RA_improved group) by T3 while 3 worsened (RA_worsened group). At the T0 baseline, however, the mean disease activity scores were similar in both groups of women (RA_improved: 3.2±0.7; RA_worsened: 3.5±1.4, p=0.7). When gene expression profiles of each RA subset was compared to the healthy women at T0, 94 genes were differentially expressed (q< 0.05; FC≥2) between the RA_improved and healthy women. These included CAMP, LGALS2, MAOA, S100A8, S100A9, S100A12 and a set of type I interferon inducible genes, many of which overlapped with previously reported RA expression signatures. A large proportion of the 94 differentially expressed genes clustered within 2 functional networks, one containing the over-expressed genes and the other containing the under-expressed genes. However, when the RA_worsened women were compared to the healthy women at T0, a largely different RA-associated expression signature was identified. Of 83 genes that were differentially expressed, only 13 overlapped with those differentially expressed between RA_improved and healthy women at T0. The majority of the 83 genes did not appear to be enriched in any relevant biological pathways.

Conclusion: In our pilot dataset, the RA-associated gene expression signatures identified before pregnancy correlated with, and were thus predictive of, subsequent improvement or worsening during pregnancy, even though mean disease activity scores at the pre-pregnancy baseline were similar between the two RA subsets. These different expression signatures suggest that there may be inherent genomic differences between women with RA that dictate how pregnancy can alter disease activity.

« Back to ACR/ARP Annual Meeting 2019

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-pre-pregnancy-rheumatoid-arthritis-gene-expression-signature-correlates-with-improvement-or-worsening-of-disease-activity-during-pregnancy-a-pilot-study/