Stillbirth: mechanisms and causes

9 Dec

This blog is a posting of the introduction section to Stillbirth, that I will publish as a page, but I wanted to have it on-line now, also as a useful introduction to the proposal for a study of stillbirth:

         A pathologist performing an autopsy is not looking for a single cause of death, but rather a chain of causation in which one event logically based on experimental and observational data links to another until a final cause of death is identified. Any terminology for the parts of that chain or overall classification system usually become tangled in these chains of causation. 

         Perhaps the most clearly defined term is the immediate or final cause of death. In utero this is relatively simple. Death is considered the irreversible cessation of the fetal heart pumping. The fetal-placenta-maternal system can keep the fetus alive despite absent function of many organ systems such as lungs, kidney, brain, etc. as is often demonstrated by infants born alive without those organs. The components needed for life are those needed for receiving oxygen, and for removing respiratory and metabolic acids. These systems include the blood, the heart pump, adequate patent vessels from fetus to umbilical cord to placenta, and a viable maternal circulation of the placenta. Anything that interferes with this system can cause death.

         We usually distinguish multiple mechanisms that are immediate causes of death, that is they stop cardiac function. These mechanisms are asphyxia, which is the interference with respiratory gases, shock which is the failure of adequate forward pumping of blood to support the needs of tissues, and heart failure in which the heart cannot fully empty, creating a back pressure which in utero is always elevates systemic venous pressure. These are useful concepts taught routinely to physicians, but clearly, they overlap, and even assigning such a mechanism as the immediate cause of death is not straight forward. For example, asphyxia causes hypoxia which increases systemic acidosis from anaerobic metabolism. The acidosis weakens cardiac contractions which then fails to pump blood forward, and develops failure with increasing umbilical venous pressure. The latter decreases umbilical venous flow and with failing cardiac forward flow increases hypoxia from decreased placental perfusion. Other starting points such a primary heart failure, for example from ventricular tachycardia, or hypovolemic shock, feed into the same final mechanisms as asphyxia. Assigning an immediate cause of death still requires a weighing of events. The pathologist, from the autopsy, is looking for anatomic evidence to clarify the order of events, the timing, and relative contribution of different factors that ended death, more than assigning an immediate cause of death*. 

         Another approach to classifying cause of death is to identify the initiating etiology. This can seem simple enough. Take the well-known example of Rh sensitization as a primary etiology. However, even in this straight-forward example, this etiology is not a complete description of all we would want to know about the fetal death. There is a complex chain of causation from sensitization to transfer of antibody to hemolysis of fetal Rh+ cells to timing of hypoxic heart failure and death. Understanding of each part of the chain has improved our therapies. The cause of death that matters depends on our objectives. If we want to look at failed prevention of Rh sensitization we may want to tabulate deaths from those who did not receive therapy because Rh was not screened for due to poverty or lack of medical resources, or those who were identified but did not receive therapeutic Rhogam on time, or we might want to look at those who became sensitized despite optimum prevention strategies, or we might want to look at infant who received in utero transfusion but died anyway or from complications of the procedure. The autopsy can provide some information on the extent of hemolysis, evidence of the severity of heart failure/hydrops, or identify therapeutic accidents. In some circumstances it may be important that the autopsy identify the cause of fetal hydrops as hemolysis from Rh sensitization.

         There are two points from this example. One is that the definition of cause may depend on the information needed to inform therapeutic practices or prevention practices, not just the biology. The second point is that the autopsy can identify some links in the chain of the autopsy from anatomy, but the investigation of death often requires a broader input.

In the case of Rh sensitization, the chain of causation is well understood. In many fetal deaths, understanding the chain is incomplete. 

         Another instructive example of a primary etiology of death is premature separation of the placenta, which may or may not be accompanied by a clinical placental abruption. The pathologist needs to evaluate the placental evidence in terms of how much placenta was acutely separated. If nearly complete, then the mechanism of death was primarily from asphyxia. If less of the placenta, but more than half, is separated of an otherwise normal placenta, then the mechanism of death is a progressive hypoxia/acidosis/ heart failure. The pathologist can estimate the timing of the separation by changes in the infarcted placental villi above the separation. If the separation is smaller than 50%, it could still be a contributing factor to immediate cause of death if the placenta already showed evidence of compromise of function, such as villous evidence of adaptation to utero-placental ischemia, of multiple placental infarctions, or severe loss of functioning perfused placentas. In such a situation, then the small separation may have caused sufficient hypoxia to be a final lethal asphyxia or to create a slower, progressive positive feedback of hypoxia/acidosis.

         Finding a placental separation and the chain of events leading to death is not the only goal of the autopsy. Discovering temporally proximal causes that initiated the placental separation are also important data. For example, evidence of fetal trauma and maternal history would point to mechanical forces separating the placenta, often an automobile accident. In other cases, there may be evidence clinically and pathologically in the placenta of preeclampsia/eclampsia in which case hemorrhage from spiral arteries is a suspected cause. There may be anatomic or clinical evidence of less well-established causes such as maternal coagulopathies, or vena caval compression. In some cases, placental separations have an increased recurrence risk unrelated to known causal factors, perhaps due to genetic differences in spiral artery remodeling or decidual trophoblast interactions. 

         Even looking for proximal causes does not completely define the chain of causation in premature placental separation. There are still many unanswered questions about the steps in pathogenesis of premature placental separation prior to and during the actual separation. If preeclampsia, coagulopathy, or recurrent/unexplained separation is present, are there anatomic predisposing reasons for the occurrence in the particular pregnancy with those risk factors. The pathologist looks for anatomic evidence of unusual changes in uninvolved basal or even parietal decidual spiral arteries that might explain a tendency to rupture and cause a hematoma that will separate the placenta. The fetal membranes are examined for venous engorgement that might indicate venous outflow obstruction. The autopsy also may yield other useful information such as the interval from the time of the separation, estimated from the timing of the overlying infarction, to the timing of intrauterine postmortem retention, to provide a estimate of the window of opportunity for rescuing the infant. The autopsy may show evidence for clues to the risk of separation, such as the presence of multiple placental infarctions. 

         I could present similar arguments for many topics including other maternal immune disease, fetal hemorrhage, maternal diabetes, etc., but these are all topics that I anticipate covering in other pages of this site. 

         My suggestion is, rather than look for classifications of stillbirth, to develop studies focused on questions about particular links in the chain of causation in fetal death. This would require many different studies and would require large multi-institution collaboration. The autopsy and placental examination can make a valuable contribution to such studies, but that the investigation of stillbirth, and of any death, requires a broad approach beyond pathological findings. The pathologist needs to be part of a team of obstetricians, radiologists, scientists and other care givers, to optimize investigations that will improve our understanding and prevention of stillbirth.  

         However, for epidemiologic purposes, investigators have sought classification schemes for perinatal death and more specifically for stillbirth. An innovative scheme was that provided by the NIH funded SCRU study1. The published paper also has an excellent critical review of classification systems. The SCRU approach recorded detailed findings, both clinical and anatomic, in an attempt at standardization. What it does not do is attempt to quantify and link the various findings into a plausible chain of events leading to the fetal demise.

(For the purposes of gathering data for the project proposed on this web site, I think these data sheets would be useful, as well as point of comparison to the SCRU studies. I have not asked for permission to publish them on the web site or use them in the study, but that is the intent. I confirmed with the NIH that Dr. Reddy has not modified the forms. The stillbirth study I am proposing would still ask detailed questions about events around the time of fetal death relating to changes in fetal movements, how and where mother slept, her diet, her perception of any illness, activities, and any medications, prescribed or over the counter, that she was taking.)

*(Different rules may apply in classification of causes in forensic/legal cases)

1.      Dudley DJ, Goldenberg R, Conway D, et al. A new system for determining the causes of stillbirth. Obstet Gynecol 2010;116:254-60.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: