Progestogen safety: implications for meta‐analysis

The meta-analysis by Sotiriadis et al.1 regarding progestogen safety is a very informative, much appreciated synthesis of trial data. The findings validate similar conclusions for efficacy by Romero et al., with reduction of respiratory distress syndrome (RDS) after exposure to vaginal progesterone for a short cervix and, supporting prior observations2, 3, quantify statistically significant harms related to exposure to synthetic 17-hydroxyprogesterone caproate (17-OHPC) in multiple gestations. The Food and Drug Administration (FDA) in the USA dealt with the safety of different progestogens in pregnancy in 1999 when addressing teratogenicity and opined that it is ‘… inappropriate to consider these drugs a single class…’, thereby suggesting these agents should be interrogated individually for safety assessments4. This approach appears optimal for current safety concerns as differing molecular structures, pharmacokinetics and pharmacodynamics of the progestogens utilized for preterm birth prevention have been demonstrated. There is ample evidence documenting that 17-OHPC is not progesterone or a prodrug of any natural progestogen and that the interaction with progesterone receptors differs between these drugs5, 6. 17-OHPC appears to act as a partial agonist for these receptors not a full agonist for particular responses7. When discussing harm related to any intervention, an essential consideration is biologic plausibility for the mechanism8. Sotiriadis et al.1 opt not to present an assessment for biologic plausibility of harm related to exposure to these agents. Their investigation suggested that this class of drugs was associated with perinatal death, RDS and composite adverse outcome in multiple gestations. However, devising a cogent argument for harm related to supplemental administration of natural progesterone to explain these outcomes is exceedingly difficult. The placental production of progesterone differs markedly during gestation, allowing human pregnancy itself to act as an informative setting for safety/toxicity. For example, if the progesterone bioavailability resulting from placental production at term or near term in a singleton or multiple gestation alone is considered safe and if this bioavailability exceeds combined placental and supplemental progesterone exposure in the midtrimester, how can dosing this natural hormone be considered harmful? Also, the apparent lack of a harmful effect related to greater bioavailability of natural progesterone at any particular gestational age in cases of higher order multiple compared with twin compared with singleton pregnancies raises additional questions regarding biologic plausibility for harm. In their meta-analysis, Sotiriadis et al.1 found that when progestogens were analyzed individually, harm was not identified for each agent and was only found related to the synthetic drug in multiple gestations. In support, to date, no Phase III study in singleton or multiple pregnancies has reported a statistically significant association for harm related to natural progesterone exposure. The authors may wish to propose considerations for biologic plausibility related to their findings of harm in multiple gestations or suggest limitations for this type of analysis. Given observations documenting a possible benefit for natural progesterone therapy from Phase III trials enrolling twins9, 10 (Table 1), safety analyses of these compounds as a class rather than as individual drugs may cast too dark a shadow on this therapeutic strategy, perhaps unnecessarily limiting investigation into populations of multiple pregnancies that could derive some benefit, and may unnecessarily force us to relive history. In summary, progestogen safety should include a consideration of biologic plausibility for mechanism of harm when concerns are identified. The analysis should optimally be based on the individual agent rather than the class of drug. Such a strategy will allow obstetricians to fully explore the capabilities of supplemental progesterone to enhance the function of the cervix, decidua, fetal membranes or myometrium in certain vulnerable populations to improve pregnancy outcomes. (1) J. M. O'Brien was involved in studies of progesterone gel treatment for preterm birth prevention sponsored by a maker of progesterone gel. (2) J. M. O'Brien serves on advisory boards and as a consultant for Watson Pharmaceuticals, a company with a financial interest in marketing vaginal progesterone gel for preterm birth prevention. (3) J. M. O'Brien and others are listed in a patent on the use of progesterone compounds to prevent preterm birth. (US Patent Number 7,884,093: Progesterone for the Treatment and Prevention of Spontaneous Preterm Birth). J. M. O'Brien*, * Division of Maternal Fetal Medicine, University of Kentucky, 800 Rose Street, Lexington, KY, USA