Reversible constriction of the fetal ductus arteriosus after maternal use of topical diclofenac and methyl salicylate

A 33-year-old physician in her second pregnancy underwent fetal echocardiography at 35 weeks' gestation. She decided to present at this late gestational age due to anxiety. Cardiac anatomy was normal, as were color Doppler velocimetry assessments of the foramen ovale, mitral valve, pulmonary artery and aortic arch. Slight tricuspid regurgitation was observed (Figure 1) and systolic and diastolic ductal velocities were 2.2 m/s and 0.42 m/s, respectively (Figure 2), indicating ductal constriction (normal values being < 1.40 m/s and 0.35 m/s). The patient did not smoke, had no history of preterm labor requiring prostaglandin inhibitors or steroids and denied use of any medication throughout her pregnancy. However, she did mention that for the last two nights, due to severe musculoskeletal pain, she had been massaging her shoulders and neck with a diclofenac gel (Cataflam Emulgel @ Novartis, São Paulo, Brazil). After rubbing in the gel she covered the painful area overnight with an adhesive patch containing methyl salicylate, l-menthol and dl-camphor (Salonpas @ Hisamitsu, São Paulo, Brazil). On the second night she repeated both topical treatments and also took a 50-mg tablet of tramadol (Tramal @ Searle, São Paulo, Brazil), a centrally acting synthetic analgesic, analogous to codeine. All medication was self-prescribed and purchased by the patient, without the knowledge of her obstetrician. She was advised to discontinue all medication and return within a week. Five days later repeat echocardiography (Figure 3) revealed normal ductal velocities (systolic, 1.1 m/s; diastolic, 0.27 m/s). An uneventful delivery occurred 1 month later and, after a neonatal echocardiogram showed no abnormalities, the infant was discharged on the 3rd day. Echocardiography at 35 weeks' gestation revealed slight tricuspid regurgitation (TR). RA, right atrium; RV, right ventricle. Echocardiography at 35 weeks' gestation measured systolic and diastolic ductal velocities at 2.2 m/s and 0.42 m/s. C ART, ductus arteriosus. Repeat echocardiography 5 days after that in Figures 1 and 2 revealed normal systolic and diastolic ductal velocities (1.1 m/s and 0.27 m/s). C ART, ductus arteriosus. The fetal ductus arteriosus is patent throughout gestation due to low oxygen tension and circulating prostaglandins. This patency allows the majority of right ventricular output to bypass the high-resistance pulmonary vessels. Rare cases of spontaneous ductus arteriosus constriction have been reported and some have been attributed to an anatomical distortion of the ductal arch2. More frequently, ductal constriction is caused by maternal use of non-steroidal anti-inflammatory drugs (NSAIDs) that enter the fetal circulation, block cyclo-oxygenase enzymes and inhibit prostaglandin synthesis. Fetuses become more susceptible to NSAIDs with advancing gestational age. Individual susceptibility also influences fetal response. Demandt et al.3 reported a case in which only one twin developed ductal constriction after maternal exposure to indomethacin. Although specific NSAID potency, dose and duration of exposure are important, cases of ductal constriction after the ingestion of a single dose have been reported4. Although oral ingestion of diclofenac, a potent NSAID, has previously been associated with ductal constriction, we did not find any reports involving topical use of the drug. Most topical medication has negligible systemic absorption and few side effects. However, there have been cases of digestive hemorrhage attributed to topical diclofenac5. Intact versus disrupted skin, age, hydration, occlusion, temperature and anatomical site all influence percutaneous absorption of medication. Improving local hydration through the use of an occlusive ointment or through physical occlusion, such as the patient did in our case, can enhance drug absorption. Skin permeability is also increased in anatomical areas with a thinner stratum corneum, such as the shoulder and back. Higher skin temperature, further enhancing topical absorption, is observed in normal pregnancy. The medium used is also important and substances suspended in liquids are known to be absorbed more readily. Gels, microscopic pockets of liquid suspended in a polymer mesh, have greater potential for penetration than creams, lotions, ointments or pastes. Some substances, such as diethyl sulfoxide, urea and propylene glycol (present in Cataflam Emulgel) are known to increase percutaneous penetration of drugs. Methyl salicylate (oil of wintergreen), a natural product obtained from a subtropical tree (Cinnamomum camphora), can also be produced synthetically and is used due to its anesthetic and antipruritic properties. Percutaneous absorption in humans is intense6 and the association with l-menthol and dl-camphor enhances skin penetration7. It is first absorbed into the bloodstream and subsequently distributed to the deeper tissues. Bleeding manifestations and elevated blood salicylate levels have been reported8 in patients who used warfarin and a topical methyl salicylate ointment. Although no studies on human pregnancies could be found, an animal study indicated that topical maternal exposure to methyl salicylate produced the same teratogenic effects as oral administration9. Our patient also used tramadol. Despite its high placental transfer rate tramadol has been used extensively for labor analgesia and has never been associated with ductal constriction. Based on the available pharmacological information, it is reasonable to assume that diclofenac and methyl salicylate applied to the skin may have had their absorption enhanced by the use of physical occlusion and the higher skin temperature typical of pregnancy. Due to their easy transfer across the placenta10 it is conceivable that they reached the fetal circulation and produced prostaglandin inhibition that resulted in ductal constriction. Their causative role is further reinforced by the fact that ductal Doppler velocimetry returned to normal after the mother discontinued treatment with topical medication. Reopening of the ductus is possible after cessation of NSAID administration1. In this case, it is impossible to determine which of the two drugs was predominantly responsible for the ductal constriction or if both of them were involved equally. This case alerts us about the potential hazards of using topical diclofenac and methyl salicylate during pregnancy. Since a variety of NSAID gels can now be obtained over-the-counter, physicians should warn their patients about their use during pregnancy. This case should also prompt physicians to ask their patient about the use of topical medication when facing a diagnosis of fetal ductus arteriosus constriction. M. R. Torloni*, E. Cordioli*, M. M. Zamith , W. J. Hisaba*, L. M. M. Nardozza*, R. M. Santana*, A. F. Moron*, * Departments of Obstetrics, São Paulo Federal University-Paulista Medical School (UNIFESP-EPM), São Paulo, Brazil, Departments of Pediatrics/Cardiology, São Paulo Federal University-Paulista Medical School (UNIFESP-EPM), São Paulo, Brazil