Archive | Uncategorized RSS feed for this section

Stillbirth Research Idea

16 Aug

I have posted the outline/synopsis of a research idea for studying the underlying causes of stillbirth as part of an intense postmortem analysis. As I have argued on this site, I believe intrauterine asphyxia is an important cause of stillbirth even in cases where a specific cause of such asphyxia had not been discovered. I think some hidden causes may be revealed if we could image the umbilical cord in utero prior to delivery of the infant. I have argued how cord wrapping can produce a functionally short cord and in one of my own daughters I have seen the effect of occult cord prolapse. The key advance that makes this study novel is the ability of MRI to identify the anatomic position of the cord in utero. I hope that a study of stillbirth will allow early identification of potentially dangerous cord entanglements and positions that could lead to novel therapies of prevention. I also hope that such information will aid in a better understanding of intrauterine asphyxia in labor and prevent complications and prevent Cesarean sections by again novel approaches to the problem.

Of course the etiologies of stillbirth are more complex and may be a confluence of factors that decrease fetal oxygenation or acid/base balance such as maternal sleep position or sleep apnea, maternal ketosis, various causes of compromised utero-placental circulation, uterine contractors, and unknown factors. This proposal suggests even more detailed postmortem history of the parents, a kind of review of systems, and narrative of events. I also propose detailed analysis of each case with some novel approached to the autopsy and placental examination to be discussed. There are also new blood markers that may be able to aid in identifying conditions such as sleep apnea after the fact. Finally, I propose the review of each case after  the postmortem evaluation is deemed complete by a team with the aim of continuing modification of the study methodology.

This study is an extension of the postmortem examination. Numerous risk factors have been identified for stillbirth, but as an example, obesity is a risk factors, but the direct mechanism of death in an individual infant involves more than just maternal obesity. The goal in each case is to improve our ability to explain the death of the infant to the parents and to ourselves. By looking at each case in detail, not for risk factors, but for a plausible and, as much as possible, provable, sequence of biological mechanisms. The constant review of the cases will accumulate knowledge that cannot be obtained otherwise, and hopefully will set the stage for controlled trials of new approaches. Stay tuned.

If anyone is interested in joining the study please email me,  I have a Powerpoint that explains some more detail. The study is still at an early point in development, and is not funded. However, commitments to the concept by obstetricians, pathologists, radiologists, and nurses will help provide concrete evidence to potential funders that the project is viable.

Open letter to obstetricians

18 Jul

Some of the pages on this blog were from a sequential series of historical chapters on fetal asphyxia starting from Dr. Little and progressing to Dr. Ron Myers key discoveries in free range monkeys, and then to later refinements in ovine models. The final chapter was on the medico-legal implications.

I have summarized this effort in the open letter that I just posted. The scientific basis of this letter is in the other posted pages on the site which review the published literature with references. I would be grateful to pathologists who utilized this site, if they would give this letter to obstetrician colleagues, and hopefully provide me with some feedback.

My email address is



Bob Bendon, MD

Postmortem retention etc.

11 Jul

My older blog had my power point lectures on the perinatal autopsy. When I gave those lectures, I had a detailed handout for each slide. I had intended to publish these as an appendix to the book that will never be published. I had not planned to blog these lectures again. However, I had promised a local pathologist who had lent me the use of a microscope to send him the references to the articles by Genest and colleagues that I had found very useful. To accomplish that goal,  I have posted a page on this topic. I hope it is useful to others as well.

Breus Mole etc.

30 May

I have added another page on subchorionic lesions on the placental surface. This was a segment of a chapter I wrote (unpublished) that followed the section on retroplacental hematoma. The previous page on placenta accreta precedes that section. A condensed versions of the RPH section will I hope soon be published in “Pathology of the Placenta: A Practical Guide” editors Khong, Mooney, Nikkels, Morgan and Gordijn.

Placenta Accreta post

20 May

This post was part of the chapter on retroplacental hematoma. I used the portion on RPH for the consensus textbook on placenta to be published. I decided to publish this portion on placenta accreta here. There are some new references to be added, which I may get to later. I feel the portion is still valuable since it reflects my experience.

I have given up trying to find a position. The position in perinatal pathology at Feinberg has been posted three times. I applied the first 2 and was neither interviewed nor my references contacted. Time to stop. However, I am grateful to those of you who do visit the site, and hope it is helpful. I have more to post if anyone finds it useful.

Posting on SNF

20 Apr

I am working on transferring blogs/pages from my old web site to this site. For SNF, I added images to the original blog. I wish I had been able to return to the slides and take new images, but I think those I have are sufficient for the main points of the section.

I had hoped to better evaluate SNF in the PROOF study, but we did not have enough cases to generate meaningful results. A collaborative study would be needed.

SNF is important to the pathologist as a clue that there may be a blood borne infection (STORCH), but there are usually other simpler ways to make such a diagnosis.

The unanswered question is why is there an association with stillbirth. Could this have been due to a selection bias, or the inclusion of more STORCH cases in the studies? Or is there a set of microorganisms that are more likely to invade the cord and cause lethal infection in the infant? In the cases I have done, no known bacterial pathogens have not been cultured from the fetus. Does SNF cause umbilical vascular injury and interfere with normal cord perfusion?

There are still some questions to answer.

MFI page added

2 Apr

Perivillous fibrinoid, I think, can best be understood as the consequence of an injury to the syncytiotrophoblast, whether ischemic, immune or metabolic, that spares the underlying cytotrophoblast. The now freed cytotrophblast do what they do at the placental margins, they secrete matrix composed of annexins, fetal fibronectin etc. These substances, along with exposed villous stroma, induce thrombogenesis from the maternal blood. Together the fibrin deposition and matrix produce perivillous fibrinoid. Diffuse syncytial injury of this type results in massive perivillous fibrinoid deposition, aka maternal floor infarction or to keep the MFI, massive fibrinoid infiltration. The diagnostic problem is discovering the various causes of this syncytial injury.


I developed this concept some years ago on my blog. When I was in Dublin, I think I heard that others have also come to a similar understanding, probably independently from me. I have now posted that older page on MFI with copious illustrations that I think supports this view of perivillous fibrinoid deposition. I added the figures within the text, rather than at the end, because there were so many.

Short case reports

14 Mar

The “pages” that I post, review my interpretation of the published literature on a topic. Like most of us, my viewpoint is skewed by the cases I have personally experienced. I illustrate the web site “pages” with illustrations from my cases. Some of these figures evolved into short case reports that I had intended to use as addendums to the “pages”. Instead, I have given each short case report its own “page” in the hopes that it alone might tempt a reader. I have just posted three more of these reports.

As I clean up old computer files, old conference presentations, orphaned photographs, etc. I find material that doesn’t seem to merit even a short case report. I no longer have access to the microscope slides, sometimes not even to all the gross images, and no access to the clinical chart. Yet, the pieces I have still present a coherent story even if incomplete. I will pass some of these along in the blogs.

I just came across 3 microphotographs. The case presents the more usual inconclusive findings from the autopsy of a stillbirth. Now, thanks to the work of Mana Parast, Theonia Boyd and others we have one more, small piece of evidence (chorionic thrombi, FTV) that the fetal death was due to asphyxia. (Even verifying this observation sometimes requires a judgment to distinguish postmortem fetal vascular changes from pre-mortem.)

Yet, in these cases of presumed umbilical cord blood flow occlusion, we are usually missing the most critical piece of information. Where was the cord at the time of death? If only we could interest obstetricians in answering this question before delivering.


The case:

Gross Diagnosis:

  1. 1. Appropriate weight male infant for 26 weeks of gestation with evidence for 2-3 days of postmortem intrauterine retention.
  2. Mild pleural and percardial effusions
  3. No malformations nor evidence of infection
  4. Fetal thrombotic vasculopathy, pre-mortem (Fig 1,2)

    Fig 1) These vilil show classic findings of HEV with an occluded lumen and extrusion of red cells and fragments.


    This stem villus demonstrates only recent in growth of endothelial cells from post mortem retention.


Discussion: The mechanism of death was most consistent with subacute asphyxia evidenced by pleural and pericardial effusions. This mechanism is likely related to the clinical history of umbilical cord entanglement. Entanglement may cause compromise of fetal umbilical blood flow by either causing the cord to become fixed in an area where compression can occur or as demonstrated in vitro, by causing a functional short cord due to the start of the cord wrapping near the umbilical cord insertion (J Matern Fetal Neonatal Med (2014) 27:1462).

The placenta did not show a specific cause of asphyxia. There were no fetal vascular thrombi, but there was regional variability in the timing of FTV lesions (hemorrhagic endovasculosis , avascular villi) that suggested a pre-mortem interference with umbilical cord blood flow (Pediatr Dev Pathol (2012) 15:275).

Short cord due to cord wrapping

4 Mar

One summer, I had a medical student for a project to study umbilical cord blood flow occlusion in vitro. We started with a simple set up to perfuse the umbilical cord from an elevated tank (Fig 1). We let the liquid flow in the umbilical vein and then wrapped it in various ways around a baby doll. Wrapping and twisting did stop flow. We switched to PVC pipes of different size (Fig 2). Fortunately the medical student was bright and persistent, and noted that the key factor was the length of the cord that was left after wrapping. However, I thought that we were just proving the obvious. The cord wrapping prevented twisting of the cord in the wrapped portion. The external twist was confined to the short, leftover segment. This result seemed to prove the obvious, if you twist a short piece of the cord, the twist per length of cord is high, compared to the twist over a long segment of cord. The student presented his research, but I did not think at the time it was worth trying to get published.

However, over time I realized that obstetricians were not aware of the possible significance of a short length of cord from placenta to the beginning of a wrap. They certainly were aware of the problem as “tight nuchal cord” at delivery, but not the potential for harm prior to delivery. I was persuaded to publish this by Michael Ross who reviewed and helped edit the manuscript for publication(1). This was many years after the study and sadly deprived the medical student of a publication when it might have been of more value to him. Still, I think that it is an important idea. Unfortunately it has not gotten much traction with obstetricians.

This study opened my thinking to the idea that in most cases of abnormal fetal heart rate tracing we do not know the precise anatomic cause. If we did, we might be able to see that some anatomical umbilical cord causes were going to be intractable and lead to asphyxia and that other might be ameliorated in some way. My experience with stillbirths also suggested that we were not discovering the cause of lethal asphyxia in many infants.

At the Star Legacy meeting in Minneapolis last summer Jason Collins, who has been trying to interest obstetricians in the umbilical cord during his career, showed MRI images of the fetus in situ that revealed detailed anatomic information about the cord (wrapping, position in relation to the pelvis, etc.). Such images in stillborn infants before delivery are a potential window to test hypotheses about cord compromise. Of course, studies of fetal death would lack Doppler of umbilical flow, but they could demonstrate a preponderance of short cord distances from placenta to the start of a cord wrapping, or occult prolapse between the presenting part and the pelvis. The study would not be easy to achieve with approaching parents in the midst of tragedy for permission for the study, recruiting a large delivery population(s) to have enough cases, and getting detailed pathological examination of the cord and placenta informed by the images.

Are there any obstetricians (beside Jason) out there that agree that this study should be done?



  1. Bendon RW, Brown SP, Ross MG. In vitro umbilical cord wrapping and torsion: possible cause of umbilical blood flow occlusion. J Matern Fetal Neonatal Med. 2014;27(14):1462-4.



Fig 1: Our equipment


Fig 2: PVC pipes

Multiple Placental Infarctions

24 Feb

I have just posted as page on this site, a section on infarctions from my old web site. It asks and tries to answer a series of questions about placental infarctions*. I posted it because I found I was re-asking myself the same questions about infarctions as I pondered the excess of multiple (>2) placental infarctions in SGA stillborn infants in my autopsy database. In most of the cases, the infarctions approached but did not exceed 50% of the parenchyma, the limit that seems to be lethal in clinical abruption (Fig 1a, 1b). I was not able to evaluate in retrospect whether the areas away from the infarctions were more severely affected by uteroplacental ischemia, than in infants with multiple infarctions who did not die, or in small for gestational age stillbirths without placental infarctions. In some cases, the photographed placental slices were mixed with small retroplacental hematomas (RPH) that acutely devitalized larger volumes of placenta than the usual infarction. A perhaps instructive case was that of a fetal death, followed the next day by lethal maternal organ infarctions(1). The infant’s placenta had multiple contemporaneous, still red, infarctions (Fig 2). They involved less than 50% of the placenta. Since the mother had appeared well before the fetus died, the acute effects of the infarctions were not likely related to decreased maternal perfusion. In some of the small for gestation fetuses, the placental infarctions could exceed 50% of the volume, but they had occurred at different points in time. We don’t know the immediate effects of an infarction on the fetus but likely it depends on how quickly fetal blood flow through the dead/dying villi is occluded, as well as the overall adequacy of placental perfusion. The important question is whether ultrasound detectable infarctions can be evaluated as a separate risk factor for death in fetal growth restriction in order to prevent fetal death. For placental pathologists, the important question is how to best make our placental examinations contribute information that will help obstetricians understand fetal risks with infarction, and prevent fetal death.

Most studies of stillbirth use relatively conventional protocols for examination of the stillborn placenta and autopsy. I think these protocols are too restrictive, and that the approach should be to do “whatever it takes” to understand the full mechanism of death. The exam should sample enough spiral artery domains to gauge the overall arterial flow by quantifying villous evidence of adaptation to decreased flow. We need to sample enough spiral arteries still embedded in the basal decidua to look for patterns of disease. We need the best estimates of volume of infarctions and how they affected fetal flow and exchange over time. We need to be able to compare one placenta to another in terms of its resistance to fetal hypoxia/ acidosis for a given a degree of maternal hypoxia or uterine contraction. I don’t mean that we know how to do this now. I am trying to define a goal. I believe that goal will require synthesizing a great deal of basic science and clinical knowledge, and also a more extensive approach to placental measurement and histological sampling.

For stillbirth, postmortem fetal changes in maternal blood flow and fetal blood flow need to be considered. The best placental samples for my purpose are those with short intrauterine postmortem retention, and in placentas from infants with fetal growth restriction due to utero-placental ischemia. Normal placentas would also be needed.


Below is an imaginary approach to a thorough placental examination. For me, it is a start to thinking about a more informative approach. I would be delighted with any thoughts or experiences from readers.


The exam would start with a photograph of the fetal surface with a 1 cm grid underlay and marking of a number of marginal points needed to align the placenta to the grid. Keeping the same orientation, the maternal surface is flipped matching the marginal points and photographed. The external points can be marked with staples, suture or whatever works. The points need to be sufficient to keep the placenta oriented on the grid. Then the placenta is sliced in 2 cm thick slices parallel to one axis of the grid. (This could be done by hand with a sharp knife. An alternative is to use an electric meat slicer. For a study of placentas with maternal anti-cardiolipin antibody(2), a second hand electric meat slicer cut very uniform slices of unfixed placenta (Fig 3).) Each slice is oriented to its grid position and sampled a number of times for microscopy. The idea is to find spiral arteries on the maternal surface (and I admit I am not yet very good at this) and include them in the histological sample. On the same sample, the idea is to relate the spiral artery to the primary villous stems on the chorionic side. These stems (40-60 in number) are also hard to locate on term placenta, but making multiple contiguous blocks would allow some to be identified. The idea histological samples would have a known relationship to primary villous stems and spiral arteries. I assume in life, the placenta attempts to match fetal to maternal perfusion; therefore it matters what portion of a villous tree is in which spiral artery blood flow. This sampling technique keeps every slide oriented to the whole placenta, and to the maternal and fetal vasculature. Some sequential blocks would be cut from fetal to maternal surface, and others parallel to those surfaces to get a different perspective. Keeping slides oriented requires a convention for orienting the samples in cassettes and then on the slides to keep the grid orientations. This overkill is purely for research in a few placentas.

The total volume could be determined by planimetry of the uniform thickness slices. Infarctions could also be planimetrically measured with some error because of the change of size with depth. The full effect of the infarct could be captured by adequate sampling of the periphery of each infarction.

Once samples are chosen the histological analysis is key and my thoughts are still in progress. This analysis needs to be more than random morphometry**. There is still value in counting knots, cytotrophoblast, vascular profiles, etc. in terminal villi, but we might be able to factor in distance from a spiral artery or level in the villous tree. Counting villous syncytial membranes is likely a good measure of oxygen carrying capacity. The work of Mayhew and that group suggest that the distance between the fetal and maternal blood flow is the most critical factor in oxygen transfer, related in to an inverse third power diffusion equation. Another possible approach is to project the outlines of villous structures onto a cellular automata grid, then using red of oxygenated, blue of for deoxygenated blood, and having the maternal at time 0 red, and the fetal blood as blue, then running a program that tries to capture diffusion and flow, the color of fetal blood in the stem veins after so may iterations may be a good measure of placental respiratory capacity for a given region. Of course, if immunochemistry could provide even more information using markers for cell cycle, growth factors, etc.




* The list of new questions, that were really old.

1: How/why do infarctions form

Usual Virchow triad in spiral arteries?

Unique transition from endothelium to trophoblast?

Where in the artery are they initiated?

Separate from small RPH or marginal separations

2: What happens in the placenta to prevent continuous fetal perfusion of the infarction?

3: What effect does an infarction have on the fetus?

Separate or with general UPI?

Timing and extent?

Is it worth detecting them, MSAFP, U/S?

Meaning of observation of multiple infarctions in SB ?

Can they be prevented?


** I am not a fan of morphometry, and not just because it is tedious. Random sampling of structured biological samples can be difficult to achieve. I attended a course in morphometry and read at least some of the two monographs of Weibel that I owned. I just don’t think the random sampling of rocks is a sound basis for understanding biological features. Admittedly some of the geometric interpretations were clever, and certainly not meaningless, but they seemed to give an incomplete picture of the biological environment.





  1. Bendon RW, Wilson J, Getahun B, van der Bel-Kahn J. A maternal death due to thrombotic disease associated with anticardiolipin antibody. Arch Pathol Lab Med. 1987;111(4):370-2.
  2. Bendon RW, Hayden LE, Hurtubise PE, Getahun B, Siddiqi TA, Glueck HI, et al. Prenatal screening for anticardiolipin antibody. Am J Perinatol. 1990;7(3):245-50.


Fig 1a

fig 1b

Fig 1b


Fig 2: Multiple acute infarctions

blog 3