Effects of conventional NSAIDs, Selective COX-2 and COX-1 inhibitors on premature labor and closure of the Ductus Arteriosus in Mice
The leading cause of neonatal mortality and infant morbidity is premature labor and delivery. Delaying labor and delivery is possible with conventional NSAIDS such as indomethacin which inhibits cyclooxygenase (COX) 1 and 2. However, COX inhibition during pregnancy may cause premature closure of the ductus arteriosus (DA) in the fetus, and, paradoxically, delayed closure of the DA post-partum in the neonate. The DA is a fetal blood vessel that joins the aorta and the pulmonary artery. Patency of the DA is essential during fetal development to divert blood away from the uninflated lungs. In the post-partum period, however, rapid closure of the DA is essential for successful transition to pulmonary respiration following birth. COX inhibition is associated with premature closure of the DA in the fetus leading to pulmonary hypertension, but to prolonged patency of the DA in the neonate leading to congestive heart failure. The relative role of COX-1 versus COX-2 in these processes is unknown.
The development of selective inhibitors of COX-2 and COX-1 enabled Loftin et al (J Clin Invest, 110:549-557, 2002) to investigate the relative effects of COX 1 versus COX 2 on timing of onset of labor, as well as on the closure of the DA in fetal and neonatal mice.
Methods: Pregnant female mice were “chronically” dosed with a selective inhibitor of either COX-1 or COX-2 on days 15-18 of pregnancy. The onset of labor was timed and compared to untreated mice. Histopathology of the DA was examined in pre-term and neonatal mice. [A number of other experiments were also performed but beyond the scope of this short summary.]
Results: Neither COX-2 nor COX-1 inhibitors administered chronically during pregnancy adversely affected the maintenance of patency of the DA in the fetus. However, selective COX-2 inhibition during pregnancy was associated with failure of the DA to close post-partum in the neonate, resulting in high post-partum mortality rates in the pups. Selective COX-1 inhibition during pregnancy had no adverse effect on closure of the DA post-partum in the neonate. Furthermore, selective COX-1 inhibition during pregnancy was efficacious in delaying on the onset of premature labor (induced by lipopolysaccharide). COX-1 inhibition also delayed onset of labor in full-term pups.
Conclusion: Selective COX-1 inhibition, but not COX-2 inhibition, delays initiation of labor without the adverse effects on the DA in either the fetus or the neonate. These findings suggest that premature closure of the fetal DA, and delayed closure of the neonatal DA, that are associated with indomethacin therapy are due to inhibition of COX-2 and not COX-1. Thus, COX-2 appears to be critical to maintain patency of the DA during fetal development and for its closure post-partum.
Editorial Comment: The safety of NSAIDs and COX-2 agents during pregnancy is an issue faced frequently by pregnant women with rheumatoid arthritis, lupus and other diseases that cause arthritis. Currently, NSAIDs are prescribed in the first two trimesters of pregnancy but withheld during the third trimester due to concerns about premature closure of the DA. Whether selective COX-2 and COX-1 inhibitors would have different safety profiles has been unknown to date. These results in mice would suggest that a COX-1 selective inhibitor might be used safely during late pregnancy to delay onset of labor without adversely affecting the ductus arteriosus. However, COX-1 specific inhibitors might not be as efficacious in controlling the arthritis since much of the inflammation in arthritic joints result from overexpression of COX-2. At any rate, these results would have to be verified in human clinical trials. Currently, aspirin is the only COX-1 specific inhibitor available on the market.