Weekly Update Information - May 29, 2001


Endometrial Receptivity: Changes in Cell-Surface Morphology George Nikas,
M.D., Ph.D., Department of Obstetrics & Gynecology, Karolinska Institute,
Stockholm, Sweden

Abstract Ovulation and fertilization trigger embryonic development and
endometrial differentiation by corpus luteum progesterone production. These
two synchronous processes couple about 1 week later, when the blastocyst
begins to implant in the now receptive endometrium (implantation window).
Receptivity is a state of endometrial differentiation marked by a change in
epithelial morphology: the hairy-like cell microvilli fuse to a single
flower-like membrane projection called the "pinopode." Scanning electron
microscopy of sequential endometrial biopsies shows that pinopodes form
briefly (1-2 days), and their numbers correlate with implantation. On
average, the formation of pinopodes is earlier in stimulated (days 19-20)
and later in artificial (days 21-22) compared with natural cycles (days
20-21). There is, however, a wide (up to 5 days) variation between women in
the cycle days on which pinopodes form. These results suggest the existence
of a narrow and discrete implantation window in humans. Detection of
pinopodes is a potential clinical marker to assess endometrial receptivity.
[Sem Reproductive Med </thieme/SRE/public/archive/2000/toc-1803.html>
18(3):229-235, 2000. © 2000 Thieme Medical Publishers, Inc.]

Introduction

Implantation is an important step in establishing a pregnancy and is of
major concern in the management of infertility. Failure at this step
greatly limits the success of in vitro fertilization (IVF) treatments.
Implantation is particularly difficult to study, for it requires a
blastocyst to interact with a receptive endometrium, that is, in vivo
conditions. Indeed, our knowledge of what happens during the first week of
human life in vivo is limited to a handful of observations.[1-4] According
to assumptions based on this and other clinical data, ovulation and
fertilization occur in the oviduct on day 14 of an ideal cycle (Fig. 1). In
the next days, the zygote floats downstream, undergoing mitotic divisions,
to enter the uterine cavity at the morula stage, on day 18. On day 19 a
blastocyst is formed, which sheds the zona pellucida and on day 20 starts
to implant in the endometrium. Meanwhile, the endometrium, under the
control of steroid hormones, differentiates and reaches the state of
receptivity. The onset of mammalian implantation can be seen as a
successful meeting of two separate processes: embryo development and
endometrial differentiation. A synchrony between these functions is
important, thus defining a bifactorial, transient period when implantation
can start, called the implantation or nidation window.[5,6]

        The phenomenon of endometrial receptivity has been extensively
studied in rodents. In the rat during the 12-hour receptive state, the
ultrastructure of the luminal epithelium undergoes striking modifications.
The apical membranes of the secretory cells lose their microvilli and
develop large ectoplasmic protrusions.[7] These protrusions were found to
be involved in pinocytosis and were thereafter termed pinopodes.[8] There
is abundant experimental evidence that pinopodes provide a specific marker
for uterine receptivity in rats. For instance, pinopodes are present only
at the time of blastocyst attachment,[9] and they follow displacement of
receptivity with antiprogestins.[10] Structures resembling pino-podes
present at the time of implantation have been described in all mammals
studied so far, including humans. Because of the obvious importance of the
endometrial factor, the formation of pinopodes has been studied with the
aim of developing a specific marker for uterine receptivity in clinical
practice.

The Surface Morphology of the Human Endometrium The endometrial epithelium
consists of two types of cells that are easily distinguishable by scanning
electron microscopy (SEM): the secretory and the ciliated cells. The
morphology of ciliated cells does not change much during the cycle. In
contrast, the secretory cells bear microvilli (MV) and undergo
hormone-dependent changes.[11,12] The temporal changes of cell surface
ultrastructure in the human endometrium have been investigated by SEM
throughout the menstrual cycle. In these studies, up to four sequential
endometrial biopsy specimens were taken from the same individuals at
48-hour intervals from natural cycles, controlled ovarian hyperstimulation
(COH) cycles for IVF, and hormone-controlled (HC) cycles with
administration of estradiol and progesterone. Participating patients were
either fertile volunteers or patients enrolled in IVF or egg donation
cycles, and ethical approval has always been obtained.

Natural Cycles Some consistent morphological SEM features occurring during
the natural cycle are as follows.[13,14] In the proliferative phase the
cells vary greatly in size and their shapes are either elongated or
polygonal. Cell bulging is minimal; the intercellular clefts are barely
marked and the MV are sparse, becoming more numerous towards the late
proliferative phase. During the secretory phase the morphological changes
are distinct and allow dating of the tissue in a 24- to 48-hour interval.
Taking an ideal 28-day cycle as reference, an increase in MV density and
length is noticed on days 15 and 16; and the cells begin to bulge, mainly
at the central part of their surface. On day 17, bulging increases,
involving the entire cell apex, and the microvilli reach their maximum
development, being long, thick, and upright (Fig. 2). On day 18, the MV
start to diminish in size and their tips may appear swollen. On day 19,
there is a pronounced and generalized cell bulging. The MV decrease further
in number and length by fusing together or disappear. Smooth and slender
membrane projections begin to form, arising from the entire cell apex
(developing pinopodes). By day 20 the MV are virtually absent, and now the
membranes protrude and fold maximally (fully developed pinopodes). Fully
developed pinopodes assume many shapes resembling flowers or mushrooms
(Fig. 3). On day 21, bulging decreases and small tips of MV reappear on the
membranes, which are now wrinkled, and the cell size starts to increase
(regressing pinopodes). By day 22, the pinopodes have virtually
disappeared, and the MV have became more numerous. Day 23 is characterized
by a further increase in the size of cells, which by day 24 begin to appear
dome shaped and covered with short, stubby MV. By day 26, the cell
membranes appear degenerate and devoid of MV. It should be emphasized that
these changes refer to an ideal 28-day cycle. In reality, although the
sequence of changes is unvarying, the actual cycle days when these changes
occur including pinopode formation may vary up to 4 days between women.
Fully developed pinopodes have been detected on days LH+6 to LH+9 (days 19
to 22) in different individuals. These individual variations in the timing
of pinopode formation are of clinical interest and will be discussed
further.

Scanning electron micrograph of endometrial epithelium on day LH + 7 of a
natural cycle. Most secretory cells bear fully developed pinopodes, which
may protrude beyond the length of the ciliated cells.      

Controlled Ovarian Hyperstimulation Cycles The effects of COH on
endometrial surface ultrastructure have been examined in oocyte donors
undergoing IVF cycles.[15] The stimulation protocol included
gonadotropin-releasing hormone a (GnRH-a) beginning in the midluteal phase
of the preceding menstrual cycle. Follicular stimulation with human
menopausal gonadotropin and follicle-stimulating hormone is initiated on
day 3 of the menstrual cycle and continued until administration of human
chorionic gonadotropin. From each donor, two to four endometrial biopsy
specimens were taken at 48-hour intervals between days 14 and 24 of the
stimulated cycle (oocyte aspiration designated day 14). The results showed
that endometrial morphology is similar to that seen in natural cycles.
Again, fully developed pinopodes were detected in only one sample from each
donor, indicating a short life span. The cycle day on which pinopodes
formed varied between women within a range of 5 days, from day 18 to day
22. In the majority of cases, pinopodes were already formed by day 19,
which was significantly accelerated for an average of 1 to 2 days in
comparison with natural cycles. Moreover, accelerated pinopode formation
correlated strongly with preovulatory progesterone rise (>6 ng by day
13).[16]

Hormone-Controlled Cycles A large number of HC cycles have been examined
with a variety of medications. In general, the hormone protocol included
pituitary down-regulation with a GnRH-a in cycling patients. The
proliferative phase was induced by a fixed or incremental dose of oral
estrogen for 1 to 3 weeks. Then vaginal or intramuscular progesterone was
added. The day of progesterone start was designated day 15 (P1). Two or
more biopsy specimens were taken from each patient, between days P6 and
P10. The surface endometrial morphology was found to be similar to that of
normal or COH cycles. The cycle day on which pinopodes formed varied
between women within a range of 3 days, from P6 to P8, in most
regimens.[17] However, in cycling patients not receiving GnRH-a, these
interpatient variations extended from P6 to P10.[18]

In one study, the correlation between pinopodes and pregnancy was
investigated in 17 mock HC cycles preceding transfer of donated
embryos.[19] Pinopodes were scored according to their number in three
grades, abundant, moderate, or few, depending on the percentage of the
endometrial surface occupied by pinopodes (>50%, 20-50%, and <20%,
respectively). Following an identical transfer cycle, all five patients
with abundant pinopodes became pregnant, three out of seven with moderate
pinopodes became pregnant, and none of the five patients with sparse or no
pinopodes became pregnant.

In another study, the clinical utility of pinopodes for prediction of the
implantation window on an individual basis was explored.[20] Candidate
embryo recipients go through a mock HC cycle with biopsy on cycle days 20
and 22 (P6 and P8) for examination by SEM. A diagnosis is made on the basis
of the following parameters: (1) number of pinopodes present and their
grades scored as already described and (2) stage of pinopode development:
developing, fully developed, or regressing. The most receptive day of the
cycle corresponds to fully developed pinopodes or is postulated to be 1 day
before regressing or 1 day after developing pinopodes are observed. A
transfer cycle follows in which synchronization with the embryo is arranged
so that the predicted most receptive day coincides with embryonic age day 6
(Fig. 4). It is assumed that by that time the IVF embryo is ready to
implant. Results of this study encourage the use of SEM in monitoring
endometrial differentiation and timing of embryo transfer on an individual
basis.

Synchronization between donated/frozen eggs and the endometrium of the
recipient. The receptive day is estimated according to a preceding
assessment cycle. DO, day of oocyte aspiration of the donor; P1, day of
progesterone start in the recipient. 

Discussion The results of these studies suggest that pinopodes are accurate
markers of the implantation window in women. Pinopodes last for less than 2
days in all cases, and the timing of their formation depends both on the
hormone treatment applied and on the patient's individual response. On
average, they form on days 20-21 in natural, days 19-20 in COH, and days
21-22 in HC cycles (Fig. 5). Such short duration and discrete timing of the
window of receptivity could significantly affect the outcome of assisted
reproduction treatments. In natural cycles, we may assume that there is an
inherent synchrony between the maturing uterus and the developing embryo,
ensuring that both will meet at the right stage. In IVF cycles, embryonic
development is probably delayed because of the in vitro conditions[21]
while the uterus may be advanced,[15,22] resulting in an early closure of
the nidation window before the zygote eventually reaches a stage capable of
initiating implantation (Fig. 6). Consequently, it would be highly
desirable if the window of receptivity in IVF cycles could be postponed for
a couple of days. This has been successfully induced in the rat following
postovulatory antiprogestin administration,[10] and a similar treatment has
been proposed for humans.[23] Indeed, a pilot study showed that
administration of a low dose of mifepristone on days 14 and 15 caused
delayed pinopode formation.[24] It is interesting that the number of
pinopodes varied between patients, some showing plentiful and others only
sparse pinopodes, with a strong correlation between pinopode number and
implantation success after embryo transfer in a subsequent cycle.[19] These
data argue positively for the relevance of pinopodes to implantation and
also that menstrual cycles in the same individual are similar. There is
some direct evidence for the latter as well, deriving from observations on
two repeated cycles studied in the same individuals (n = 5) under the same
hormone regimen. Not surprisingly, the surface endometrial morphology
during the second cycle was found to be similar to that in the first cycle
(G. Nikas et al, unpublished). Some women show very few or no pinopodes
despite regular ovulation and menstruation. Such cases are usually
associated with abnormalities in epithelial morphology, including the
presence of large hyperplastic cells or dense microvilli arranged in tufts.
It is possible that abnormal endometrial maturation might be a cause of
infertility. Thus, examination of endometrial biopsy specimens by SEM for
pinopodes may be a useful test in infertility work-up.

The cellular and molecular function of pinopodes in humans remains unknown.
As mentioned before, tracer experiments in the rat and mouse[8] have shown
that these structures perform pinocytosis. However, similar experiments in
the baboon failed to reveal any pinocytotic function (A. Fazleabas,
personal communication, 2000). This may be true for the human as well
according to some preliminary results (G. Nikas et al, unpublished). In any
case, the smooth surface of the pinopodes and the accompanying loosening of
epithelial cell-to-cell contacts[13,24] may facilitate trophoblast adhesion
and/or penetration to the stroma. In an in vitro model using endometrial
epithelial cells growing on stromal cells, human blastocysts adhered only
on areas bearing pinopodes.[25]

The size of pinopodes in SEM is around 5 mm, arising from the entire cell
apex. Considering the shrinkage of the tissue because of dehydration in SEM
preparation, their native size may be around 8 mm. With this size,
pinopodes are visible in conventional histology as bulbous protrusions of
the cell apices.[26] However, the ability of histology to assess pinopodes
is limited by the low resolution of light microscopy and by the fact that
histology is performed on tissue sections, where only a small area of the
surface can be examined. Thus, it appears that SEM is indispensable for
detecting and scoring endometrial pinopodes.

In conclusion, the human implantation window, as determined by the
formation of uterine pinopodes, is short, and its timing varies with
different hormonal treatments and in different individuals. Examination of
endometrial biopsy specimens for pinopodes is a potential test in
infertility evaluation both for the optimization of embryo transfer and for
supporting other studies on human implantation, which remains the "last
frontier" in infertility treatment.

Acknowledgements: Studies described in this article were performed in
collaboration with the following institutions: Alexandra Maternity
Hospital, Athens; Human Reproduction Centre, Athens; Infertility Unit of
the Hammersmith Hospital, London; Instituto Valenciano de Infertilidad,
Valencia; The Family Federation of Finland, Helsinki; and The Jones
Institute of Reproductive Medicine, Norfolk, VA.

Reprint requests: Dr. Nikas, Department of Obstetrics & Gynecology,
Karolinska Institute, Huddinge Hospital K57, S-18164 Stockholm, Sweden.

Pregnancy Inadvisable for Some Women After Peripartum Cardiomyopathy

WESTPORT, CT (Reuters Health) May 23 - The health risks associated with
subsequent pregnancies among women with a history of peripartum
cardiomyopathy can include significantly decreased left ventricular
function and death, according to a report in the New England Journal of
Medicine for May 24.

"Patients with a history of peripartum cardiomyopathy with persistent left
ventricular dysfunction have a high incidence of developing symptomatic
heart failure during subsequent pregnancy," Dr. Uri Elkayam, from the
University of Southern California School of Medicine, Los Angeles, told
Reuters Health. He added that in a recent study, 19% of such patients died
after subsequent pregnancy.

Dr. Elkayam and colleagues reviewed medical records of 44 women who had had
peripartum cardiomyopathy and a total of 60 subsequent pregnancies. In
addition, the researchers conducted interviews with the women and their
physicians.

Among the 16 women with persistent left ventricular dysfunction, the
incidence of therapeutic abortion was high, Dr. Elkayam noted. He also
pointed out that women in this group who did not end their pregnancy had a
higher incidence of premature delivery.

"Therefore, because of the risks to the mother and especially because of
the increased possibility of death, we advise against pregnancy in these
women. Since patients who had a therapeutic abortion did better than
patients who did not, women who conceive after peripartum cardiomyopathy
who continue to have left ventricular dysfunction may consider an
abortion," Dr. Elkayam told Reuters Health.

Of the 28 women who had become asymptomatic, 21% developed symptoms of
heart failure during subsequent pregnancies. However, mortality in this
group was zero, Dr. Elkayam stressed.

"Women with peripartum cardiomyopathy who have recovered left ventricular
function need to know that there is some risk to their health in subsequent
pregnancy. However, the risk of death is very small," Dr. Elkayam said.
"Based on these findings, women can make their own decisions about
pregnancy, based on a much better understanding of the risk of
complications."

N Engl J Med 2001;344:1567-1571,1629-1630.