COMMENTARY Perinatal deaths after sildenafil treatment of fetal growth restriction raise the issue of safety in randomised clinical trials Niccolò Lombardi 1 | Giada Crescioli 1 | Alessandra Bettiol 1 | Claudia Ravaldi 2 | Alfredo Vannacci 1,2 1 Department of Neurosciences, Psychology, Drug Research, and Child Health, Section of Pharmacology and Toxicology, University of Florence, Tuscan Regional Centre of Pharmacovigilance, Florence, Italy 2 CiaoLapo, Charity for Healthy Pregnancy, Stillbirth, and Perinatal Loss Support, Prato, Italy Correspondence A. Vannacci, Department of Neurosciences, Psychology, Drug Research, and Child Health, Section of Pharmacology and Toxicology, University of Florence, Viale G. Pieraccini, Florence 6‐50139, Italy. Email: alfredo.vannacci@unifi.it Sildenafil citrate (SC) is a drug developed for male erectile dysfunction (ED), specifically able to block phosphodiesterase (PDE) type 5, a nitric oxide‐cyclic guanosine monophosphate (cGMP) catabolising enzyme found in vascular smooth muscle. 1 SC is under evaluation for several medical uses, such as Raynaud's phenomenon, stroke, achalasia, endo- thelial dysfunction, and neonatal‐ and adult‐onset pulmonary hyper- tension (PH). SC has been also experimented in the management of fetal growth restriction (FGR), a major cause of perinatal mortality and morbidity. 2 Some pregnancies with FGR have in fact elevated peripheral maternal vascular resistance in uterine arteries, and SC seems to increase uterine blood flow and to potentiate oestrogen‐ induced vasodilation. A 2017 systematic review evaluating maternal tolerance and obstetric and perinatal outcomes following SC use in human pregnancy concluded, despite limited data, that there does not appear to be any severe adverse maternal side effects, nor increase in stillbirths, neonatal deaths, or congenital anomalies rates attributed to SC. 1 Recently, five multicentre randomised placebo‐controlled trials have been launched (STRIDER trials), 3 and some results have already been published. 2 In contrast with the evidences reported in the above mentioned review, authors concluded that SC did not prolong preg- nancy or improve pregnancy outcomes in severe early‐onset FGR, and no relevant safety issue was reported. Unfortunately, the attention on SC trials raised globally when a Dutch trial using SC to correct FGR has been halted after 11 babies died unexpectedly. 4,5 A review by a committee of the STRIDER trial declared that lung complications were more common in the babies born to women given SC, of whom 17 had lung problems, and 11 died, while only three babies in the placebo group developed lung complications, and none of them died, suggesting a potential causal association. To date, no evidences have been published on the possible role of SC on new‐borns' lung complications. However, a possible biological mechanism of toxicity could be found in SC withdrawal after birth, with a worsening of babies' pulmonary function through a rebound effect. This phenomenon is known in Pharmacovigilance as “Type E” (end of use or withdrawal) adverse drug reaction, 6 typical of drugs able to induce long‐term modifications of cellular metabolism and subse- quent adaptations of body homeostasis. In patients with PH, sudden cessation of vasodilator agents, such as SC, carries a significant and unpredictable risk of rapid clinical deterioration, 7 with pulmonary com- plications beyond what expected from the simple cessation of active treatment. Although no report of this kind exists regarding sildenafil withdrawn in new‐borns, birth itself obviously represents an abrupt treatment cessation. This could be particularly relevant since pre‐ existent pulmonary dysfunctions are common in FGR neonates, who are also prone to develop lung complications after birth, such as perinatal asphyxia, meconium aspiration, and persistent PH. 8 No other trial to date reported such an association, but that should not be considered a proof of safety. Randomised studies are often inadequate to evaluate drug safety, because only frequent and acute adverse events may be assessed, whereas unknown, rare, and/or long‐term latency ones are difficult to be detected. 9 This is of further relevance in paediatrics, since children have complex devel- opmental and pharmacological characteristics that vary from adults: Relying on adult safety and efficacy data when dealing with children can have unpredictable and tragic effects. 10 In this particular case, SC was shown to reduce mortality and improve oxygenation in neonates affected by PH, 11 but clinical deterioration after sildenafil cessation was also a known issue in adult patients with PH. 7 The paper is an original work not previously published in whole or in part and is not under consideration for publication elsewhere. Received: 9 November 2018 Accepted: 30 December 2018 DOI: 10.1002/pds.4740 Pharmacoepidemiol Drug Saf. 2019;1–2. © 2019 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pds 1