In Depth

Case Studies from the Wisconsin Stillbirth Service Program - #2

Richard M. Pauli, M.D., Ph.D.

This is the second in an indefinite series of presentations of specific cases referred to the Wisconsin Stillbirth Service Program that we think are instructive and illustrative. This time I have chosen instances which are good examples of the importance of internal postmortem evaluation (formal autopsy) in arriving at a specific diagnosis.

Introduction

The formal internal postmortem evaluation is an integral part of comprehensive assessment following stillbirth. In the sequential series of stillborn infants assessed through WiSSP, 33.9% of all autopsies demonstrated one or more fetal abnormalities. Of course, some of these are trivial or incidental findings.

However, we have also tried to calculate how frequently internal postmortem findings are crucial in demonstrating an etiologic cause of a baby’s death. We did this using a sort of reconstruction: as each case was reviewed, the question was asked, “If this one segment of the evaluation had not been completed, would diagnosis nonetheless have been made?” Each time the answer was “no”—that portion of the evaluation was judged to be diagnostically critical for that case. Results of such a stepwise reconstruction showed that internal autopsy findings played such a crucial role in about one-fourth of all stillborns assessed (and in about 60% of those in whom a specific diagnosis was found).

Because in many instances more than one evaluative step had been judged to be critical, and in order to compare the relative importance of each portion of the protocol, these data were transformed into “diagnostic equivalents”. For each case, each critical evaluation was divided by the total number of evaluations judged to be critical in that instance. So, for example, if in a given instance photographs and radiographs are both judged to be crucial in generating a particular etiologic diagnosis, then each would be assigned a diagnostic equivalent equal to 0.5. So, at least in a relative sense, these values reflect the numbers of diagnoses which would have been missed had that portion of the protocol never been completed. Once again, the internal postmortem had the highest yield. The estimate of diagnostic equivalents was 18% for the internal postmortem examination. That is, in 18% of referrals a diagnosis would not have been made had an internal examination not been completed.

Despite such demonstration, some still advocate selective internal autopsy. That is, some suggest that it not be completed on severely macerated fetuses, or that if external morphologic evaluation is normal then it not be completed, or if a diagnosis is obvious by examination alone that it need not be done. The three case studies discussed here illustrate how the autopsy can reveal unexpected results or can provide information critical to confirmation of a suspected diagnosis.

Example 1: A female infant was delivered at about 32 weeks gestation. She was mildly macerated but otherwise had what appeared to be a single malformation. The right arm was clearly anomalous (Fig. 1) with apparent absence of the radius on that side as well as absence of the thumb and deformity of the second finger. [Because there was a positive family history of a cousin with an isolated limb deficiency birth defect, one might have been tempted to speculate that these two events were related.] Radiographs confirmed the absence of the right radius and of any ossific centers of the thumb; the second metacarpal was very thin and the bones of the second finger hypoplastic (Fig. 2).

In addition, however, those x-rays showed hemivertebrae of the sixth and seventh thoracic vertebrae (Fig. 3). Chromosome studies showed a normal, female karyotype. Internal postmortem assessment showed, additionally, absence of the right kidney and hemiagenesis of the uterus.

Diagnosis and comment:  The presence of the three major malformations identified in this baby (radial ray deficiency, vertebral malformation, renal agenesis) means that she had what is termed the VATER (or VACTERR) association. These are acronyms for malformations that are known to co-occur at a frequency greater than expected by chance: Vertebral anomalies; Anal atresia; Cardiac defects; Tracheo-Esophageal fistula; Radial ray anomalies; Renal hypoplasia or agenesis. It is generally agreed that any baby with at least three of these features can be diagnosed with the association. Obviously, internal postmortem assessment was here crucial in uncovering the third of the necessary three characteristics that allowed for a definitive diagnosis to be made. It ended up that the family history was irrelevant (a different kind of limb deficiency); fortunately that ‘easy’ explanation did not result in truncation of the evaluation protocol.

We still do not know the reason that the particular birth defects of the VATER association tend to co-occur. However, what has become clear from experience with many families is that VATER association almost never recurs. So, establishing this diagnosis allows for completely reassuring counseling — essentially no risk for recurrence of the VATER association and, assuming that intrauterine death arose as a result of this process, probably no significant increased risk for recurrence of stillbirth either.

Example 2: This baby was delivered at 35 weeks gestation and, while liveborn, died within the first 5 minutes following delivery. Photographs showed no clearly abnormal characteristics (Fig. 4): there was more hairiness than average of the face; the ears were border-line low set and over-folded. Radio-graphs were normal. Tissues were not obtained for cytogenetic evaluation. Because of vaginal bleeding at around 30 weeks, there was suspicion that this baby’s death might be attributable to abruption of the placenta. Placental pathology showed only one small area suggestive of a small, previous abruption. Internal autopsy was normal except for the left lung. It was clearly enlarged and pushed the contents of the mediastinum to the right. It weighed three times its normal value. The upper lobe contained a multicystic structure which histologically was found to have all of the features of what is termed Cystic Adenomatoid Malformation of the Lung.

Comment: Cystic adenomatoid malformation of the lung is a hamartomatous lesion (a benign, abnormal development of embryonic tissue). Prior to its ultrasonographic diagnosis, as many as 60% of infants with this birth defect were either stillborn or died in the immediate newborn period. Often it is associated with hydrops, presumably because of pressure on the vessels leaving the heart and consequent changes in blood flow and/or heart failure. It appears to be a non-genetic process. Had this infant not had internal postmortem assessment, all one could have told the family is that no cause had been identified for their daughter’s death. This diagnosis, at least, provided an answer to why she had died and allowed reassurance to be given that they had no significant increased risk for untoward outcomes in subsequent pregnancies.

More recently, with the advent of relatively routine ultrasound evaluations, many fetuses with cystic adenomatoid malformation of the lung are identified prenatally. It has become one example of sometimes successful intrauterine surgery. When cystic adenomatoid malformation of the lung of the type seen here (microcystic or type III in the Stocker classification) is identified, intrauterine surgery to completely remove the lobe of lung that is involved can result in a healthy baby being born.

Example 3: This infant was the product of a pregnancy that ended at 35 weeks gestation following preterm labor and absence of fetal heart tones on presentation at the birthing hospital. External examination showed no particularly unusual features (with a comment only that the lobules of the ears were upturned); indeed the photographs obtained show subtle unusual facial features (Fig. 5),

but certainly nothing that would suggest a specific diagnosis. Radiographs were normal. Likewise cytogenetic studies showed a normal female chromosome makeup. Internal postmortem examination, however, showed that this baby had renal and ureteral agenesis on one side and a small, dysplastic kidney on the other.

Diagnosis and comment: Had a hierarchical protocol been followed, likely no internal postmortem would have been performed. Its completion did allow for a diagnosis of renal adysplasia. Renal adysplasia can variably cause abnormal growth of one kidney, absence of one kidney, abnormal growth of one and absence of the other, abnormal growth of both or absence of both. Clearly, simple absence of one kidney usually has no health consequence, while absence of both is uniformly lethal because of secondary pulmonary hypoplasia.

Renal adysplasia sometimes is familial, caused by a variably expressed single autosomal dominant gene. Therefore, family evaluations were carried out which showed that this baby’s father had unilateral renal agenesis. Thus, the final diagnosis in the family is Hereditary Renal Adysplasia. This diagnosis — an important one for the family — would not have been discovered if the normal appearance of this fetus had led to the conclusion that internal postmortem assessment was unnecessary. Once made, it implies high risk for transmission of the gene that causes it (50%) and a substantially increased risk for lethal malformations to recur.