Chronic inflammatory disease and its treatment during pregnancy

Abstract

The decision to have children is often coupled with varying degrees of apprehension. Women with chronic disease often worry about how the disease itself or its treatment will affect pregnancy and the fetus. The aim of this thesis was to add to the current knowledge concerning pregnancy and birth outcomes in chronic inflammatory disease.

First, we studied pregnancy and delivery complications in women with Crohn’s disease and ulcerative colitis, the main types of inflammatory bowel disease (IBD). By using register data, we constructed a proxy model for disease activity, categorizing women according to (i) no activity, (ii) stable disease and (iii) flaring disease. Compared to women without IBD, women with Crohn’s disease and ulcerative colitis more often delivered by emergency and elective cesarean sections. Women with ulcerative colitis had a higher risk of venous thromboembolism during pregnancy, which seemed to be the highest in women with flaring disease.

In the second study, we assessed risks of adverse birth outcomes in women with Crohn’s disease and ulcerative colitis and their infants, by diagnosis and by the disease activity proxy model. Compared to women without IBD, women with Crohn’s disease and ulcerative colitis more often gave birth preterm, particularly those with flaring disease and with thiopurine treatment during pregnancy. Small for gestational age and stillbirth were more common for women with Crohn’s disease, and were also related to flaring disease.

The third study focused on anti-tumor necrosis factor (anti-TNF) treatment and birth defects, a relationship which has not been well-studied. Anti-TNF is used to treat IBD and other chronic inflammatory disease such as rheumatoid arthritis, ankylosing spondylitis, psoriasis and psoriatic arthritis. We included all infants to women with these diseases in Sweden and Denmark and determined their anti-TNF treatment status in early pregnancy. Infants of the general population were included for a frame of reference for descriptive purposes, while formal comparisons were limited to the women with chronic inflammatory disease. The distribution of birth defects in all three groups was similar and ventricular septal defect, atrial septal defect, congenital hydronephrosis and hypospadias were the most common. Rates for corrective surgery were also similar. Comparing infants to women with chronic inflammatory disease, with and without anti-TNF treatment, the risk of any birth defect or defect of a particular organ-system was slightly but imprecisely increased, and an association could not be determined.

In the fourth study, we identified all live-born singleton cases of preterm birth as cases, and births occurring at a later gestational age as their controls, among women with IBD. We found that significant disease activity and treatment with thiopurines or anti-TNF predicted preterm birth. The study illustrated the challenges of confounding by indication. Some misclassification of disease activity was likely not completely avoided, leading to residual confounding and making it impossible to assess the effects of disease activity and immunosuppressive treatment separately. However, weighing the risks of relapsing disease against those of drug treatment, disease activity seems more important to avoid.