Retained placenta: will medical treatment ever be possible?

The standard treatment for retained placenta is manual removal whatever its subtype (adherens, trapped or partial accreta). Although medical treatment should reduce the risk of anesthetic and surgical complications, they have not been found to be effective. This may be due to the contrasting uterotonic needs of the different underlying pathologies. In placenta adherens, oxytocics have been used to contract the retro-placental myometrium. However, if injected locally through the umbilical vein, they bypass the myometrium and perfuse directly into the venous system. Intravenous injection is an alternative but exacerbates a trapped placenta. Conversely, for trapped placentas, a relaxant could help by resolving cervical constriction, but would worsen the situation for placenta adherens. This confusion over medical treatment will continue unless we can find a way to diagnose the underlying pathology. This will allow us to stop treating the retained placenta as a single entity and to deliver targeted treatments.

retained placenta

We need to stratify treatment of retained placenta according to underlying cause.

The scale of the problem

Obstetric hemorrhage remains a leading cause of maternal death worldwide 1 and retained placenta (RP) is a significant contributor 2. The incidence appears to be related to intrapartum intervention, as rates increase with both time and health system development 3. The rate currently in Europe is around 2–3%. If RP is not treated, it may lead to maternal death due to postpartum hemorrhage or sepsis. Historically, in the UK, RP led to the deaths of around seven women per 100 000 births 4, and in modern African settings, with delayed access for manual removal, the case fatality rate remains around 1% 5. It is clear, however, that many RPs will spontaneously deliver with time. Indeed, in the placebo arm of the largest randomized trial to date, 38% delivered spontaneously in the hour following recruitment, with 10% having blood loss of >1000 mL 6. Thus, there is a careful risk–benefit balance between waiting, which allows more spontaneous deliveries, and intervening to prevent blood loss and sepsis. It has been suggested that when considering PPH, the optimal time for manual removal is 18 min 7 but this does not take into consideration the adverse medical, neonatal or psychological effects of manual removal of placenta.

Herman et al. first documented the ultrasound findings in the normal third stage of labor and placenta adherens 8 and went on to describe the use of ultrasound for the diagnosis of the three known subtypes: placenta adherens, partial accreta and trapped placenta 9. However, following this description there has been no formal analysis of the utility of ultrasound in RP.

Currently, the standard treatment of RP is manual removal whatever the subtype, even though this has anesthetic and surgical risks. While treating RP with drugs would remove many of these risks, medical therapies have not been well investigated, partly because of a current inability to distinguish rapidly and accurately the various forms of RP on the labor ward. The use of medical therapies to treat RP, irrespective of the underlying subtype, has led to confusion.

Conflicting guidance

Umbilical vein injection of oxytocin

Following the hypothesis that the pathogenesis of the RP adherens is due to the failure of contraction of the retro‐placental myometrium 8, umbilical vein injection (UVI) of oxytocics to counteract that inhibition seems to be a plausible remedy. However, conflicting advice on the use of UVI oxytocin to treat the RP has been issued by the National Institute of Health and Care Excellence (NICE) group in the UK. They recommended its use in 2007 because of Cochrane Reviews that showed that it was associated with significantly fewer manual removals when compared with UVI saline 10. Now, NICE has changed its advice and no longer recommends the use of UVI oxytocin, following a cost analysis and a finding that it tends to be associated with severe postpartum hemorrhage despite being an oxytocic 11. This is in keeping with other large controlled trials, including the Release Trial 6, that evaluated UVI oxytocin and demonstrated lack of significant overall benefit. Likewise, the administration of intravenous oxytocic agents to deliver an RP is not recommended unless the woman is bleeding excessively, although there is no conclusive evidence on this.

UVI prostaglandins

There is also frustration with the use of other UVI agents. Meta‐analysis has demonstrated a significant reduction in the need for manual removal of the placenta (MROP) following UVI prostaglandin when compared with UVI oxytocin 10. Randomized controlled trials also showed that intravenous infusion of sulprostone also reduced the need for MROP when compared with placebo 12. However, there are only a few of these trials, with small sample sizes, and their effects were inconsistent or statistically insignificant. Therefore, UVI prostaglandins are currently not recommended.

Nitroglycerin

The use of nitroglycerin for the trapped placenta is also clouded with uncertainty. One trial showed sublingual nitroglycerin to be associated with a significant reduction in manual removal of the placenta, when compared with placebo 13. However, this trial involved only 24 women and a subsequent larger study by the same team showed no significant benefit 14. Another small trial of intravenous nitroglycerin also demonstrated no benefit 15. We await with interest the outcome of the UK‐wide randomized controlled trial on the use of glycerin trinitrate for RP (GOT‐IT Trial).

Why does the optimal management of RP continue to elude us?

It is likely that the various medical treatments for RP do not work for a number of reasons. The first reason is the lack of accurate diagnosis of the RP subtypes. Although RP is a group of pathologies, it has a common presentation and has been therefore treated as a single entity. It is important to get the type of RP right before prescribing uterotonics instead of relaxants. For, although uterotonics like oxytocin can treat placenta adherens by contracting the retro‐placental myometrium, they are likely to exacerbate trapped placenta by increasing cervical constriction. In contrast, relaxants can counteract uterine or cervical constriction to deliver a trapped placenta, but they would worsen placenta adherens and possibly precipitate hemorrhage by increasing uterine relaxation. Neither uterotonics nor relaxants can treat a partial accreta, but continuous cord traction on an accreta, following a false impression that a medication administered might help, could also lead to severe hemorrhage or uterine inversion.

The second reason may be that UVI agents probably go to the wrong place. After passing up the umbilical vein they perfuse through the placenta villi into the radial veins in the uterus without reaching the retro‐placental myometrium. It is only on the second pass around the body that they reach the myometrial capillaries and initiate the desired myometrial contraction. By this stage the concentration is likely to be very low and no longer directed at just the retroplacental myometrium (see Figure 1).

Figure 1

Blood flow from umbilical veins to fetal capillaries.

Conclusion

Effective medical therapy for RP should avoid the surgical and anesthetic risks associated with the current standard treatment of manual removal. However, it can only be made to work if the reasons for its failure are adequately addressed. UVI may not directly enter the myometrial capillary bed and alternative forms of directed treatment, possibly using nanoparticles, may provide a better way of delivering oxytocics to the retro‐placental myometrium. But the main problem is our inability to differentiate between the various pathologies underlying RP, each of which has a different, and conflicting, treatment requirement. The confusion over medical treatment will continue unless we stop treating the RP as a single entity and start tailoring the treatment to the underlying pathology. This will take an effective way of diagnosing the type – probably using a mixture of uterine ultrasound and Doppler of the uterine vessels. Until then, the medical treatment of RP is likely to continue to elude us.

Funding

No specific funding.