IN THE LIT

R. M. Pauli, M.D., Ph.D.

Increased midtrimester amniotic fluid activin A: a risk factor for subsequent fetal death.Petraglia F, Gomez R, Luisi S, Florio P, Tolosa JE, Stomati M, Romero R. Am j obstet gynecol 180:194-197, 1999.

This group took advantage of ‘leftover’ amniotic fluid specimens to retrospectively assess whether they could identify a marker that might predict fetal death later in pregnancy. A total of 110 amniocentesis specimens (obtained between 15 and 17 weeks gestation) were analyzed for two compounds that the authors thought might be related to death in utero: activin A and corticotropin-releasing factor. Elevation of each has been previously correlated with adverse outcomes such as preterm labor and preeclampsia. After measurement, the values were compared for the 10 pregnancies that resulted in fetal death (all occurred between 17 and 27 weeks gestation) and the 100 controls who were delivered at » 37 weeks gestation, were appropriate size for gestational age and who had no birth defects.

Corticotropin-releasing factor showed no relationship with fetal death. High activin A was related to subsequent fetal death (median of 5.9 ng/ml vs 3.0 ng/ml) although there still was quite marked overlap between the two groups. Using a stringent cutoff value (i.e. maximizing specificity for fetal death) does result in some predictive power. Of 7 samples above the 95% confidence interval for activin A level, 4 had intrauterine deaths, while of the remainder, only 6 of 113 resulted in fetal death. Or, thought of another way, a normal activin A reduces risk of fetal death from about 9% to 5% overall, while an elevated value increases that risk from 9% to around 57%.

Why does activin A correlate with fetal death? Is it that activin A is necessary for critical functions, that it increases in response to stress, or that it reflects “failing systems”? That is unknown.

It would be of considerable benefit if there were antecedent markers that would predict risk for subsequent fetal death, if such a marker told us what should be done. Until the questions of why activin A and intrauterine death are correlated are answered, this latter question will probably remain unanswered, too. In the future, activin A, either from amniotic fluid or from maternal blood, may be added to the tests that are done routinely. It is not ready for prime time yet, though.

Maternal cigarette smoking, regular use of multivitamin/mineral supplements, and risk of fetal death. The 1988 National Maternal and Infant Health Survey. Wu T, Buck G, Mendola P. Am j epidemiol 148:215-221, 1998.

Smoking during pregnancy causes a modest increased risk for stillbirth (probably around a 30% increase compared with non smokers). It is not clear whether smoking causes these excess fetal deaths because of direct effects on placental and fetal vasculature, because of decreasing availability of nutrients depleted through their use in inactivating toxins associated with tobacco use, because of direct damage of tobacco toxins, or because of biologic co-variables (e.g. relative nutritional inadequacies) in the smoking population. Given the possibility that nutrient depletion plays a role, then nutrient supplementation might provide some protective effect from maternal smoking for the developing fetus. This study used data from the 1988 National Maternal and Infant Health Survey to assess whether vitamin/mineral supplements might decrease the fetal death risk associated with maternal smoking.

Supplements did not affect fetal death rate among non-smoking women. However, the odds ratios for death dropped rather dramatically in the babies of smoking women. Obviously, vitamin and mineral supplements should not be discouraged in non-smoking women! However, these data do suggest that particular emphasis on such supplementation is critical in women who do choose to continue to smoke during a pregnancy.

Depression in women suffering perinatal loss. Carrera L, Díez-Domingo J, V Montañana V, Monleón Sancho J, Minguez J, Monleón J. Internat j gynecol obstet 62:149-153, 1998.

This is a study of depression (as measured by the Beck Depression Inventory) following perinatal death. There were three groups. The ‘treatment’ group consisted of 23 women suffering a perinatal loss who were provided what the authors call ‘psychological intervention’ (see below). In addition, there were 34 women who suffered a perinatal loss who were provided no such counseling or intervention, and 37 women who had healthy liveborns. The authors are to be credited with using such control groups. Indeed, the design of this study is (I think) better than for much of the grieving/mourning literature. I only think it is because it isn’t exactly clear what interventions were carried out in the actively treated group. The authors say that “the intervention group received a 1-year psychological support” consisting of encouraging of seeing, touching holding, naming etc; explanation of the mourning process; directed counseling to avoid pregnancy for a year to avoid the ‘replacement baby’; encouraging of expression of feelings. There is no indication how often contact occurred, whether it was or was not ongoing, who provided counsel etc. Furthermore, for the two perinatal loss groups assignment almost certainly was not random, but the authors say nothing about how these two groups came to be. I really hate articles from which I can’t actually tell what was done

Nonetheless, some of the data may be of use. First, they show that those whose baby had died had higher initial depression indices (no big surprise) that remained elevated at 6 months postpartum (still not surprising). At 12 months, the counseled group’s Beck indices had returned to those seen in the liveborn control group. In contrast the uncounseled group still had, on average, higher depression scores than the control group. This suggests that the kinds of interventions all of us think are appropriate really do help parents. Unfortunately the flaws of design make the data less than convincing.

Clinical and pathologic correlates of stillbirths in a single institution. Ogunyemi D, Jackson U, Buyske S, Risk A. Acta obstet gynecol Scand 77:722-728, 1998.

This is an assessment of 115 consecutive stillborns occurring in a single hospital (despite the journal of publication, a hospital in New Jersey). The data presented are an interesting set of lessons about stillbirth assessment and its interpretation.

First, consistent with most other studies, 23% were found to have congenital malformations (of which around 15% of the total were thought to be clearly causally linked to the death). Implausibly, however, nearly 65% were thought to have died from placental and cord related processes (37% placenta and 28% cord). These are hugely discrepant with our own experience and with most others if an appropriate level of skepticism about when findings are etiologically relevant is maintained. To their credit, the authors list the specific pathologic diagnoses. Perusing that list offers at least a partial explanation — “torsion of cord”, “congestion of cord”, “entangled cord”, “focal hematoma of cord”. Interpretation of causality, I think, was over-enthusiastic and ‘under-skeptical’.

Contribution of congenital malformations to perinatal mortality. A 10 years prospective regional study in The Netherlands. De Galan-Roosen AEM, Kuijpers JC, Meershoek APJ, van Velzen D. Eur j obstet gynecol reproduct biol 80:55-61, 1998.

Unlike the study just reviewed, this one sought only to assess the contribution of fetal malformations to the occurrence of perinatal death (both stillbirth and early infant death). Stillbirth here has a curious definition — fetal death after 28 weeks gestation or death within the first 30 minutes after delivery. Virtually complete ascertainment from one region of the Netherlands resulted in a total study population of 247 perinatal deaths, or whom 155 were “stillborn”. Apparently (this is not absolutely clear in the article) all were assessed consistently including clinical examination, internal postmortem evaluation, radiographs, bacterial and viral cultures and titers, placental assessment, and, selectively, cytogenetic evaluation. So, I would take exception in substantial ways with the character of the evaluations completed: the bacterial and viral cultures and titers were probably mostly a waste of time; chromosomal evaluation used selectively has already been shown to miss critical diagnoses. Nevertheless, this is a nice study because of the virtually complete ascertainment without any bias of referral of stillborn infants in a defined geographic region.

In 25% of the stillborns congenital malformations were identified. However, the authors felt that only 13% had major (and thus likely etiologically relevant) malformations. I am uncertain why this is as low as it is — perhaps because those between 20 and 28 weeks gestation are not included; perhaps because of missed diagnoses; perhaps at least in part because some of those listed under ‘minor’ or ‘non-lethal’ malformations are, in fact, well associated with a marked increase risk of stillbirth (e.g. thanatophoric dysplasia).