Non-steroidal anti-inflammatory drugs are not fully safe for fetus: comments on the article Treating common ear problems in pregnancy: what is safe? by Vlastarakos et al.

Abstract

Sir, I read with great interest an article by Vlastarakos et al. [1] entitled Treating common ear problems in pregnancy: what is safe? The paper was written by clinicians based on information collected from various databases. However, certain interpretations and conclusions regarding non-steroidal anti-inXammatory drugs (NSAIDs) should be clariWed from the teratological point of view. It is true that speciWc NSAIDs, presently known as nonselective cyclooxygenase (COX) inhibitors, have not proven to be human teratogens with the exception of high doses of aspirin [2–5]. However, in a recently published paper by Ofori et al. [5]––conducted on a large (n = 36,387) representative population––a greater risk for all the structural congenital anomalies (OR = 2.21; CI = 1.72–2.85), especially cardiac septal defects (OR = 3.34; CI = 1.87–5.98), was revealed due to the use of NSAIDs. This article is the only fully available document that presents human prenatal toxicity of selective COX-2 inhibitors (COXIBs). It should also be noted that according to Kozer et al. [4] (OR = 2.37; CI = 1.44–3.88) and Martinez-Frias et al. [6] (OR = 3.47), a signiWcantly higher risk of gastroschisis was highly dependent on the use of aspirin along with maternal age and smoking. The above mentioned results were conWrmed by animal data presented in various meta-analyses that showed a higher incidence of ventricular septal defects (VSD), celosomy (gastroschisis, umbilical hernia), and diaphragmatic hernia [7–9]. Additionally, in the case of congenital VSD in rats, an insigniWcantly higher ratio was calculated for ibuprofen (106.5/ 10,000 exposed fetuses) and aspirin (46.26/10,000) when compared with various control groups (4.56–19.72/10,000) [9]. For celosomy, it was found that only aspirin (56.41/ 10,000), as a single drug, signiWcantly increased the risk of abdominal wall defects that in untreated rat populations ranged 0.3–1.54/10,000 [7, 8]. It is worth mentioning that a signiWcant decrease of fetal body weight was noted in COX-inhibitor-exposed groups when compared with corresponding controls [8]. The above Wnding carries importance, because intrauterine growth retardation is one of the most common signs of prenatal toxicity [10], even though it was not revealed in human epidemiological studies for NSAIDs [11]. It is also unclear why there is no information about risk of miscarriage in the article. Epidemiological, retrospective studies showed that NSAIDs increased incidence of fetal death and the results depended on the time of treatment [11]. The risk increased from 1.26 (CI = 0.85–1.87) to 6.99 (CI = 2.75–17.74) when the drugs were used for 10– 12 weeks or just a week before miscarriage, respectively. Even with all the limitation to the study by Nielsen et al. [11], the results are important since they were proved later by Li et al. [12]. The risk was much higher when drugs were used around conception (OR = 5.6; CI = 2.3–13.7) or longer than 1 week (OR = 8.1; CI = 2.8–23.4). Other complications, like the ductus arteriosus constriction with secondary persistent pulmonary hypertension and reduced renal perfusion that led to oligohydramions, were pointed by Vlastarakos et al. [1]. F. Burdan (&) Experimental Teratology Unit of the Human Anatomy Department, Medical University of Lublin, 4 Jaczewskiego Str, 20090 Lublin, Poland e-mail: fb3@wp.pl 1 odds ratio 2 95% conWdence interval