You all probably saw the sexy articles that were presented at the European Society for Human Reproduction and Embrylogy meeting in Prague last week. You know, the ones about the clowns during embryo transfer, and the one about how 20 anovulatory women who got therapy started ovulating, thus proving that we all have to just relax. Well, I was curious enough to go read all the available abstracts. Below is my summary. Because I used to be a scientist I've probably used more scientific language than is helpful, although I've tried to be clear. I hope it's helpful.
There was a lot going on at this meeting. Which puts the lie to H's rather cynical comment when I told him where Dr Candour was last week: "Good to know he's off drinking vodka and visiting titty bars." Then we both cracked up at the thought of Dr Candour in a titty bar. Sadly H's comment probably says more about H's experiences with sales conferences than about how our esteemed doctors spend their time.
I should point out that lots of studies contradict each other. I focused on the talks rather than the posters (at scientific meetings the sexy stuff gets presented in person, and the less sexy or less prestigious stuff is just stuck on a wall for people to look at if they have time). I lost the will to live part way through reading the abstracts for the 1000 or so posters. I also focused on the stuff that's relevant for me, so I'm sorry, I didn't read the stuff on PCOS or anything much on male infertility. I'm fairly sure you will forgive me.
Protocols, response rates and egg quality
- Antagonist (short) protocol gives lower pregnancy rate (Al-Inany, Egypt)
- Antagonist protocol doesn't give lower egg quality in a subsequent FET - pregnancy rates with blastocyst transfer in FET cycles are the same whether blastocysts came from antagonist or agonist cycles (Lee et al, Korea)
- The flare protocol is better than both long protocol and antagonist in poor responders (meta analysis of other studies by Franco et al, Brazil)
- In most cycles these days, the stimulation drug is recombinant FSH - rFSH (e.g., Follistim/Puregon). Making recombinant FSH involves isolating the gene and inserting it into a production cell culture like chinese hamster cells (lots of cells grown together in a nutrient broth). The hormone is purified out of the broth. Only the pure hormone is produced. This means that unlike un-medicated cycles, there is no involvement of LH in priming the follicles for release. This was not an issue with earlier protocols because earlier stimulation drugs (e.g., Menopur) were produced by purifying the urine of menopausal women, so there was a low level of LH present. This has been an area of investigation - to see if adding low levels of LH will improve the egg quality.
Adding rLH to stimulation protocols which use pure rFSH in agonist (long) cycles reduces the rate of miscarriage (pregnancy rates are the same). You might therefore expect that protocols using puregon/follistim which are pure recombinant (r)FSH, would give a slightly higher miscarriage rate than Menopur which contains LH as well in small quantities because it's produced by purification from human urine. This was not tested - the study just looked at adding rLH (e.g., Luveris) to the rFSH. Abdelmassih et al, Brazil.
- BUT adding rLH to rFSH in a long agonist protocol did not result in higher ongoing pregnancy rate. Andersen et al, Copenhagen, Denmark. My thought is that perhaps women's metabolism is so finely tuned that it's hard to see population effects here, it's about the individual woman and her response. But then I'm not an endocrinologist.
- AND Pre-treating with rLH prior to starting rFSH stimulation in a long agonist cycle (treating with the rLH for a week after down-regging) resulted in a higher percentage of embryos fertilised (number of embryos was the same). No difference in pregnancy rate. Fleming et al, Glasgow, UK and lots of other places,
- Women with an abnormal response to FSH stimulation given their age and in vivo FSH level may have FSH-receptor mutations. Lalioti et al, Yale, USA.
- Taking longer to reach the blastocyst stage leads to poorer pregnancy rates in fresh cycles, but not in frozen ones. The hypothesis is that ovarian hyperstimulation may advance the day of receptivity of the uterus, so that putting back blastocysts a day later than the usual day 5 results in lower implantation - the window of implantation has already passed. The same would not be true in a frozen cycle. Mitwally et al, Detroit and Michigan, USA.
- Prolonged treatment with DHEA does improve ovarian function in women with decreased ovarian reserve. Although a small sample size, there were significant improvements in oocytes retrieved, fertilised oocytes, and normal day 3 embryos. Gleicher and Barad, New York, USA.
- Antagonist protocol for priming egg donor recipients has a higher pregnancy rate than the traditional agonist treatment. Casan et al, Valencia, Spain.
Implantation and embryo quality
- Endometrial receptivity is governed by the upregulation and down-regulation of 100s of genes. This gene expression is different between un-treated cycles at day 7 after the LH surge when compared to ovarian hyperstimulation cycles at day 7 after hcg. At days 1, 3, 5, and 9 the gene expression pattern was the same between the two sets. This indicates that at the day of maximum receptivity the endometrium differs in a stimulated cycle, possibly leading to lower receptivity - although we don't know that. Horcajadas et al, Valencia, Spain.
- Embryo scoring parameters that affect implantation significantly are: Nuclear alignment; nucleolar (the nucleolus is one part of the cell nucleus) number, size and alignment; day 2 cell number, symmetry and nucleation. Everything else did not significantly affect implantation rate (ie many of the characteristics our clinics use including day 3 cell number and alignment etc, and the morphology at day 5). L. Scott, address not given.
- Many embryos are mosaic - i.e., some cells are chromosomally normal, while others are abnormal. Oocytes are much less frequently abnormal than the resulting embryos (7-10% versus 70% in young women), indicating that errors come during embryonic cell division, rather than arising in the egg itself. The only way currently to screen for mosaically abnormal embryos is PGS, which is controversial and has yet to be proven in a large scale study to result in higher levels of implantation. Lundin, no address given.
- Cigarette smokers have less receptive uteruses (uteri?). Statement of the bleeding obvious by Soares et al, Valencia, Spain.
Other factors affecting cycle outcome
- With a yah, boo, and sucks to the article which got all the publicity, which showed that some women with no ovulation who received therapy started ovulating, this study shows that levels of stress had no effect on IVF cycle outcome. Lintsen et al, Nijmegen, Netherlands.
- Women with premature ovarian aging have a lower pregnancy rate and higher miscarriage rate, need higher doses of stimms yada yada yada, but interestingly their embryos do NOT show a higher rate of aneuploidy. I.e. something else is causing the higher miscarriage rate , not a simple eggs are crap --> embryos have the wrong chromosomes problem. Weghofer et al, USA
- Women with recurrent fetal loss (pregnancy loss after seeing cardiac activity) had increased expression of inhibins a1β1and a4β1 compared to women with recurrent loss of empty gestational sacs. They can't figure out the mechanism. Anim-Somuah et al, Liverpool, UK
- The strongest predictive factor for secondary recurrent miscarriage (after number of previous miscarriages) is the gender of the first child - having a boy is hypothesised to activate the mother's immune system. How that fits with all the controversy around the immunology stuff I don't know. Nielsen et al, Copenhagen, Denmark
- Peripheral blood concentration of natural killer cells (PBNK) bears no relationship to uterine concentrations of natural killer (uNK) cells. PBNK levels don't vary during the menstrual cycle. Levels of PBNK were similar between recurrent miscarriers who carried their pregnancy to term versus those who had another miscarriage. Rai et al, London UK (From Lesley Regan's lab)
- Uterine NK cells have a strong role to play in implantation, providing the mechanism guiding maternal epithelial cells towards the embryonic trophoblas cells via oxygen sensing and synthesis of a compound called VEGF. Higher concentration of uNKs leads to higher pregnancy survival rate in pigs - concentration goes up three-fold at critical points in healthy embryonic development. Croy et al, Ontario, Canada.
- The egg's polar body can be used to do genetic testing even before selecting which eggs to use for insemination (very creative technique developed by an Italian group in response to the law in Italy which says that only 3 embryos can be created and all must be transferred). More than 50% of eggs in women over 35 had aneuploidy. Ferraretti et al, Bologna, Italy.
- BUT this misses post-fertilisation genetic problems. Evidence from Delhanty (not sure of address) says that more than 50% of genetic abnormalities in the embryo are post-meiosis and would therefore not be picked up in polar body testing. They found a similar ratio of abnormal embryos in those with recurrent miscarriage and those with recurrent IVF failure.
- Women with recurrent miscarriage or multiple IVF failures produce very few viable embryos. Only 7.5% of embryos surveyed were normal with PGS, only 1.5% implanted and only 1% produced ongoing pregnancies. Bianchi et al, Yale, USA.
- Severe male factor infertility contributes to a high rate of chromosomally abnormal embryos (95% in non-obstructive azoospermia). Magli et al, Bologna, Italy.
- Endometriosis may be present in everyone, and only become a pathology when a certain stage is crossed. Not clear what that stage is
- Women with mild or moderate endometrisis have a decreased ovarian reserve as measured by anti-mullerian hormone (AMH) which is produced by small antral follicles. Antral follicles are also more diverse in size in women with endometriosis compared to those without endo. Cunha-Filho et al, Porto Allegre, Brasil.
- Surgery of even mild endo has a positive effect on pregnancy and live birth rates. Mohammed Saleh et al, Saudi Arabia
Long term effects of IVF/ICSI
- ICSI children at 8 years old had no substantial differences in development to children who were not the result of ICSI. There was a higher incidence of congenital malformations (10% vs 3.3%) but these were all correctable by minor surgery.
- Vanishing twins lead to a higher incidence of intrauterine growth restriction later in pregnancy for the surviving singleton. The later the twin vanishes, the more likely there is to be an issue. Pinborg et al, Denmark