Use of non-steroidal anti-inflammatory drugs in pregnancy and oligohydramnios: a review

Abstract Objective The aim of this review is to evaluate the relationship between the use of non-steroidal anti-inflammatory drugs (NSAIDs) during last trimesters of the pregnancy and the reduction of amniotic fluid. Methods Electronic databases were searched (PubMed, Medline, and Scopus). Selection criteria included studies reporting the relationship between oligohydramnios and use of NSAID during pregnancy. We analyzed the median age of women, weeks of pregnancy at the beginning of the drug administration, kind of medication, period of exposure and dosage, deepest vertical pocket (DVP), and amniotic fluid index (AFI). Results Of the 68 records identified, we analyzed 29 studies investigating the administration of NSAIDs, including 11 studies examined the administration of the Indomethacin, four articles have focused on the use of Nimesulide, and only two manuscripts considered the use of Diclofenac. We found a strict correlation between the development of oligohydramnios and the use of NSAIDs. The oligohydramnios is reversible, and the normal amount of amniotic fluid is restored after the interruption of the treatment. Conclusions The use of NSAIDs should be considered when maternal benefits outweigh the potential fetal risk, at the lowest effective dose for shortest duration. Beyond 48 h of NSAIDs treatment, we consider ultrasound monitoring of amniotic fluid, and we suggest stopping therapy if a decline AFI is present.


Introduction
The non-steroidal anti-inflammatory drugs (NSAIDs) are a class of drug with analgesic, anti-inflammatory, and antipyretic effects [1,2].The primary mechanism of NSAIDs action is the inhibition of the activity of cyclooxygenase (COX) enzyme and a consequent reduction in prostaglandin levels.The enzyme COX, also identified as prostaglandin endoperoxide H synthase (PGHS), occurs under three isoforms: COX-1, COX-2, and COX-3 [3].Given the effectiveness and security for the general public, these drugs are widely used, even among potentially at-risk categories, such as pregnant women [4].Prevalence data of the NSAID use during pregnancy are not easy estimable because these medicines are available over the counter (OTC) and without medical prescription.Limited records about their consumption can underestimate their real use [5].Since 1975, the United States Food and Drug Administration (FDA) ruled a pregnancy risk classification related to different drug use during three trimesters [6].NSAIDs are classified in category B (there are no fetal risk on animal reproduction studies or in women in the first trimester) during first and second trimester and in category D (there is evidence of human fetal risk) for third trimester [6].Doses of aspirin >150 mg/d are considered category D during all trimesters, while to low-dose aspirin is attributed to the category C (according to animal reproduction studies, there is an adverse effect on the fetus and there are not controlled trials in humans; potential advantages could justify the use of the drug in pregnant women despite potential risks).However, several studies reported prenatal exposure to NSAIDs could cause fetal and neonatal adverse effects that also involve several organs [5].Exposure in the first trimester was related to an increased risk of miscarriage and malformation.Conversely, the use in the third trimester is associated to a premature closure of the fetal ductus arteriosus and oligohydramnios.Fetal and neonatal adverse effects affecting the brain, kidney, lung, skeleton, gastrointestinal tract, and cardiovascular system have also been reported after prenatal exposure to NSAIDs [7].In the past, FDA advised pregnant women and health care professionals to interrupt NSAID and aspirin use during the last 3 months of pregnancy from 28th week to the end of pregnancy.On 15 October 2020, FDA communicated that pregnant woman should avoid use of NSAIDs around 20 weeks or later because of the onset of oligohydramnios and kidney problems in new-borns [8].The topic of our review is to assess the correlation between NSAIDs administration in pregnancy and the development of oligohydramnios.

Protocol, eligibility criteria, information sources, and search strategy
Data from PubMed, Medline, and Scopus electronic databases were used to review existing literature about use of NSAIDs in pregnancy.We included studies from March 1988 to June 2022, utilizing combinations of the relevant medical subject heading (MeSH) terms, key words, and word variants for "pregnancy", "NSAID", and "oligohydramnios".To collect as many cases as possible, no restriction on language, publication date, or publication types was applied in the selection of articles.Studies about multiple pregnancies and animal species were ignored.Manuscripts not related to the topic of our review were excluded.Following aspects were analyzed: the median age of women, weeks of pregnancy at the beginning of the drug administration, kind of medication, period of exposure and dosage, deepest vertical pocket (DVP), and amniotic fluid index (AFI).

Study selection and data extraction
The scope of this research is the detection of mother, and fetus-outcomes following administration of NSAID.Only studies reporting and describing the relationship between oligohydramnios (AFI � 5 cm or DVP < 2 cm) and use of NSAID during pregnancy were considered eligible for the inclusion in the present review.The first assessment, carried out by two investigators (AS and DD), was supervised by a third investigator (VD).
The assessment of full texts was performed by the same investigators.To improve the diagnostic accuracy of the research, a manual assessment of the list of references was performed.If more than one study was published on the same cohort with identical endpoints, the report containing the most comprehensive information on the population was included to avoid overlapping populations.

General characteristics
A total of 68 records were identified from the searches of PubMed/Medline databases.After removing 39 studies during title and abstract screening, 29 manuscripts were selected for full-text review.More specifically, 39 studies were removed for following reasons: multiple pregnancies studies (n ¼ 7); off-topic studies (n ¼ 18); manuscripts which recruited animals (n ¼ 3); trials with full text not available (n ¼ 10); a study duplicated (n ¼ 1).Twenty-nine records, therefore, were included in the qualitative analysis.Of the 29 manuscripts, 11 were case reports, four prospective trials, two retrospective studies, nine reviews, one casecontrol, one randomized control trial, and a short communication.The median age and gestational age of drug administration of the study population ranged from 20 to 36 years and from 20 to 35 weeks, respectively.The population sizes of the studies ranged from 6 to 124 patients.Eight studies were conducted in Europe [9], eight trials in United States [10], four studies respectively in India (two articles) [11,12], Israel [13], and Argentina [14].The main features are listed in Table 1.

Synthesis of the results
A total of 29 studies investigating the administration of NSAIDs have been included in this paper.Of these, 11 studies examined the administration of the Indomethacin [10,14,16,[20][21][22][23][24][25][26][27], four articles have focused on the use of Nimesulide [15,[17][18][19], while only two manuscripts considered the use of Diclofenac [9,12].Other analyses assessed the action of drugs that have fallen into disuse during pregnancy [13].For most patients, the dose ranges from 100 mg to 200 mg for Nimesulide, from 100 mg to 300 mg for Diclofenac and from 50 mg to 200 mg for Indomethacin; the period of drug assumption was from 3 days to 56 days (Table 2).

Indomethacin, diclofenac, and oligohydramnios
The largest sample was studied by Savage et al. [14], that in 2007 enrolled 124 patients at 24 weeks of pregnancy.Of these patients, 72 women assumed 100 mg of Indomethacin, 23 assumed 50 mg, 21 received 200 mg, and five patients took 150 mg of the drug.The dosing regimen described were per day.They were treated for 21 days (6-65 days) and 7.3% developed oligohydramnios (AFI <50 mm).Despite different doses have been administered, no significant relationship was found between drug's administration, dose, and the onset of the oligohydramnios [14].In October 2013, Klauser et al. [10] published a randomized clinical trial, where they

Studies including animals
recruited 87 women taking 100 mg of Indomethacin per day between 24 and 32 weeks of pregnancy.After 21 days of assumption of the drug, the authors observed oligohydramnios in five patients (5.7%).In 1990, Hendrick et al. [24] published a case-control study, where they included 67 cases and 67 controls.Thirty-seven women were treated with Indomethacin 100 mg followed by 75 mg every 6 h and 30 patients assumed Ibuprofen 600-800 mg every 6 h for 28 days on average.Overall, 26/37 (70%) and 8/30 (27%) patients developed oligohydramnios, respectively.The reduction of amniotic fluid was reversible in cases compared to controls.Only 2/67 women had a decrease of AFI and it did not resolve before delivery.Sandruck et al. [16], in their prospective study, found 2/34 patients (6%) with 44 mm and 36 mm of AFI.Bivins et al. included 33 women at 32 weeks of pregnancy that assumed Indomethacin 150 mg daily for 43 days.Of this cluster, 13 patients (38%) developed oligohydramnios.In advanced gestational age of pregnancy, the percentage of oligohydramnios is higher than second trimester of pregnancy, but the risk of decrease of AFI is also significant at 21 weeks [22].The relationship between Diclofenac and oligohydramnios was also examined.Scherneck et al.
[9] and Phadke et al. [12] published two case reports in 2015 and 2012, respectively.In both cases, patients received Diclofenac 150-300 mg and 50 mg for about 20 days, and they developed oligohydramnios at 21 weeks of pregnancy.

Nimesulide and oligohydramnios
Sawdy et al. [17], as matter of fact, investigated the effect of another kind of NSAIDs, Nimesulide, during pregnancy.They began to administer Nimesulide to 44 women at 21 weeks of pregnancy and they continued for 52 days.Of these, 54% developed oligohydramnios even if the dosage of NSAID was not specified.Grincevi� cien _ e et al. [15], Sawdy et al. [17], and Paternoster et al. [18] published a recent case report where they investigated the correlation between the use of Nimesulide and the reduction of AFI.In all cases, the assumption of this drug caused the onset of reversible oligohydramnios (100% of patients).Although these studies and other previous reports were based on a small sample of participants, they reported similar findings [19].The use of other NSAIDs investigated, like piroxicam and magnesium dipyrone, is contraindicated in pregnancy.

Discussion
NSAIDs are commonly prescribed in pregnancy to treat fever, pain, and inflammation; these are used chronically for inflammatory bowel or chronic rheumatic diseases such as rheumatoid arthritis and spondyloarthropathy [28].During pregnancy, NSAIDs cross the placental barrier and could cause embryo-fetal and neonatal damage affecting the brain, kidney, lung, skeleton, gastrointestinal tract, and cardiovascular system.These side effects derive mainly from inhibition of prostanoid activity and from pharmacokinetic changes occurring during pregnancy.Adverse effects depend on the type, the dose and the duration of maternal therapy, the gestational age, the time to administration and delivery [7,29].Recent studies have not demonstrated any differences in neonatal outcomes when comparing the maternal administration of nonselective COX inhibitors with COX-2 inhibitors, because COX-2 is expressed in a variety of fetal tissues [30].Women who are treated with NSAIDs during early pregnancy can have an increased risk of miscarriage and congenital malformations.The association was stronger if the initial NSAIDs use was around the time of conception or if it lasted more than a week.The probable mechanism is the suppression of prostaglandin biosynthesis in the tissues of the reproductive system: this could lead to abnormal embryo implantation that predisposes to miscarriage [31].Furthermore, according to recent literature, the exposure to NSAIDs in third trimester of pregnancy is associated with several side effects such as oligohydramnios, bleeding diathesis, ileal perforation, premature closure of the ductus arteriosus, and acute or chronic renal injury [32].Nevertheless, the focus of our review is the assessment of the onset of oligohydramnios related to the administration of NSAIDs, including indomethacin, diclofenac, ibuprofen, and nimesulide.The precise mechanism of oligohydramnios, caused by indomethacin, is probably mediated by the decrease of fetal urine output, as shown by Kirshon et al. [23].Indomethacin is a potent vasoconstrictor of fetal blood vessels, it decreases renal blood flow and could lead to renal dysfunction [33][34][35].In fact, COX-1 is expressed in renal tubuli and COX-2 in renal medulla.Prostaglandins are involved in the mechanisms of renal circulation, and the blockade of prostaglandins synthesis could cause reduced renal perfusion, and consequent renal failure.Therefore, COX inhibitors could lead to reduced urine production in the fetus, with consequent oligohydramnios [30,36].Goldenberg et al. [25], and De Wit et al. [26] described cases of oligohydramnios occurring after administration of the indomethacin at doses respectively of 200 decreased at 100 mg/day and 75 mg/day between week 20 and week 28 of gestation.After interrupting therapy, the oligohydramnios regressed completely in utero and authors agreed to hypothesize that the effect of indomethacin on amniotic fluid volume may be dose related.Another risk of all antenatal NSAIDs, including COXIBs, is the constriction or premature closure of the ductus [19,27].Endothelial and smooth muscle cells of the ductus arteriosus express both COX-1 and COX-2 [30].During fetal life, the ductus arteriosus shunts the output of the right fetal heart from the pulmonary trunk directly to the descending aorta.It is essential during fetal life to shunt blood away from the high resistance pulmonary circulation, so much so that the ductal closure in utero may lead to fetal death or pulmonary hypertension.The results from previous studies are contradictory: some studies found that the effect of indomethacin on the constriction of the ductus arteriosus was related to the fetal serum concentration of this drug [22] by the contrast other trials demonstrated that the ductal constriction, oligohydramnios, and composite morbidity were not associated with duration of therapy, gestational age at start or cessation of therapy, time between dosing and delivery or dose regimen [37].In the retrospective cohort of Savage et al. [14], the multivariable logistic regression and similarly, the univariable analysis showed no relationship between dose regime and odds of constriction ductal and prolonged therapy (duration >75th percentile; v 2 test, p ¼ .10)and no relationship between dose regimen and odds of constriction (v 2 test, p ¼ .67).Other authors demonstrated that the rate of indomethacin-induced constriction of the ductus arteriosus was related to gestational age (from 0% before 27th week 27 to 61% at 31-34 weeks of pregnancy).Moreover, the peak diastolic velocity of the fetal ductus arteriosus after maternal indomethacin ingestion was constant at 25 cm/second before 27 weeks and increased between 27 and 30 weeks (a mean of 39 cm/second) [38].Another side effect of NSAIDS is kidney damage due to inhibition of COX; the hyperechogenic fetal kidney may be due to NSAIDs-induced lesions [39].
There are few published data on fetal adverse events after short-term NSAID exposure in the second trimester.A recent review showed that the short-term use of NSAIDs in the second trimester did not represent a risk for fetal adverse effects; however, AFI assessment should be recommended during the longterm use in the late second trimester [40].The FDA recommended avoiding NSAIDs in pregnant women at 20 weeks or later in pregnancy rather than the 30 weeks currently described in NSAID prescribing information [8].In Canada, the use of prescription and OTC NSAIDs are currently contraindicated in the last trimester of pregnancy (week 28 to the end of the pregnancy).These recommendations do not apply to the use of low dose (81 mg) aspirin for specific pregnancy-related conditions under the direction of a healthcare professional.
The strengths of our study included the fact that this is a complete review of the use of NSAIDs during pregnancy and the assessment of the onset of oligohydramnios related to their administration.Strengths of this review included a broad search strategy across multiple databases with minimal restrictions.However, the findings of this review are reliant on the quality of the underlying studies, some of which were conducted as retrospective analyses of routinely collected observational data.One limitation is the variation in definitions of fetal risk in terms of duration of treatment and dose regime between different studies.Furthermore, there are still few published data on the fetal outcomes and a short-term NSAID exposure in the second trimester.We present an illustrative comprehensive review about the relationship between NSAIDs administration in pregnancy and the onset of oligohydramnios.Pain therapy during pregnancy is still today a challenge and it implies prudence and attention by health care professionals and pregnant women.We found a correlation between the development of oligohydramnios and the use of NSAIDs even if the studies were not homogeneous, analyzing different type if medication at different dosage.The oligohydramnios is reversible, and the normal amount of amniotic fluid is restored after the interruption of the treatment.For these reasons and according to FDA recommendation, published in October 2020, we suggest that the use of NSAIDs, not only during the third trimester of pregnancy but also since 20th weeks, should be avoided.The use of NSAIDs should be considered when maternal benefits outweigh the potential fetal risk, at the lowest effective dose for shortest duration.Beyond 48 h of NSAIDs treatment, we consider ultrasound monitoring of amniotic fluid and we suggest stopping therapy if a decline AFI is present.

Table 1 .
Characteristics of the included trials.
DVP: deepest vertical pocket; AFI: amniotic fluid index; PROM: premature rupture of membranes.aData are represented as mean ± SD or median (range).bData are represented as mean ± SD or median (range).cData are represented as mean ± SD or median (range).dData detected in patient 1. e Data detected in patient 2. f Oligohydramnios related to asymmetrical intrauterine growth retardation (in one).THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE

Table 2 .
List of excluded studies and reason for exclusion.