L. V. Farland, K. F. Correia, L. A. Wise, P. L. Williams, E. S. Ginsburg and S. A. Missmer - P-values and reproductive health: what can clinical researchers learn from the American Statistical Association? - Hum. Reprod. (2016) 31 (11). doi: 10.1093/humrep/dew192
You should read the Editorial Commentary we publish this month. Studying it is mandatory for prospective authors, but also readers will enjoy this lucidly written commentary. That’s why it’s Free Access. Leslie Farland and colleagues wrote it, and it’s about the p-value. Her group of Boston authors – strongly supported by the editors of this journal – is urging readers and authors of Human Reproduction to change their attitude toward p-values. You are encouraged to welcome and discuss findings with a strong biological plausibility from carefully conducted studies, even if they lack statistical significance. Conversely, statistically significant outcomes that lack biological plausibility may be chance findings, e.g. in ‘Big Data’ studies, and should be viewed with skepticism.
Dyer S, Chambers GM, De Mouzon J, Nygren KG, Zegers-Hochschild F, Mansour R, Ishihara O, Banker M, Adamson GD. International Committee for Monitoring Assisted Reproductive Technologies. World report on Assisted Reproductive Technologies:2008-2009-2010. Hum Reprod 2016; doi: 10.1093/humrep/dew082
World ART figures show unexplained differences in ICSI usage
This month we publish the latest world ART figures in the ICMART World Report on Assisted Reproductive Technologies 2008-2009-2010. In the three years the report covers, almost 4.5 million ART cycles have been initiated (in the 61 countries contributing data). These resulted in more than one million babies (25.7% live birth per cycle started). Collecting and validating the data must have been a gargantuan effort and Silke Dyer and her many co-authors have to be commended for this on behalf of the whole world ART community.
Apart from giving numerical insights, the report (even more important!) provides benchmarks for ART procedures. And one of those benchmarks for me is very disturbing: notwithstanding the fact that it has been shown over and over again that IVF gives better results than ICSI in couples with non-male infertility, we perform too much ICSI: on average about twice as many ICSI cycles are initiated as IVF. The incidence of severe male infertility cannot explain this discrepancy. I was puzzled by this finding, however I was dumbfounded by the regional differences. In Asia ‘only’ 1.4 times as much ICSI as IVF is performed, in North America 2.7 as much, in South America 6 times as much, and in the Middle East even 60 (yes, sixty) times as much. In Europe, the figures vary from as low as 0.8-1.1 in the Nordic countries to 9 times as much ICSI as IVF in Spain, to a (blimey!) 25-fold ICSI excess in Poland.
It’s not like the ‘therapeutic illusion’ of old: the patient who will get better without treatment will also do so with treatment. No, this treatment hurts a couple’s chances. In the three years under review, ICSI on faulty indications has prevented at least 25,000 couples from getting pregnant.
Joyce C. Harper, Debbie Kennett and Dan Reisel, The end of donor anonymity: how genetic testing is likely to drive anonymous gamete donation out of business, Hum. Reprod. (2016) doi: 10.1093/humrep/dew065
Due to genetic testing donor anonymity does no longer exist.
Many thousands of people worldwide have been conceived with donor gametes but not all parents tell their children of their origin. Genetic testing will make this impossible. Over three million people have already used direct-to-consumer genetic testing. The rapidly increasing availability of cheaper and more detailed tests poses numerous challenges to the current practice of sperm and egg donation: 1. Whether they are donating in a country that practices anonymous donation or not, donors should be informed that their anonymity is no longer guaranteed, as they may be traced if their DNA, or that of a relative, is added to a database. 2. Donor-conceived adults who have not been informed of their status may find out that they are donor-conceived. 3. Parents using donor conception need to be fully informed that their children’s DNA will identify that they are not the biological parents and they should be encouraged to disclose the use of donor gametes to their children. All parties concerned must be aware that, in 2016, donor anonymity has ceased to exist.
Xavier Santamaria, Sergio Cabanillas, Irene Cervello, Cristina Arbona, Francisco Raga, Jaime Ferro, Julio Palmero, Jose Remohı, Antonio Pellicer, and Carlos Simon, Autologous cell therapy with CD1331 bone marrow-derived stem cells for refractory Asherman’s syndrome and endometrial atrophy: a pilot cohort study, Hum. Reprod. (2016) doi: 10.1093/humrep/dew042
Asherman’s syndrome is an almost incurable condition, characterized by the presence of intrauterine adhesions and endometrial atrophy. It is iatrogenic and results in menstruation disorders and infertility. Many therapies have been attempted for these conditions, but none have proven really effective. Xavier Santamaria, Sergio Cabanillas and colleagues from Valencia, Spain, have addressed the question whether cell therapy using autologous peripheral blood CD133+ bone marrow-derived stem cells (BMDSCs) could perhaps offer a safe and efficient therapeutic approach for patients with refractory Asherman’s syndrome and/or endometrial atrophy and a wish to conceive? They found that in 16 patients, in the first three months following treatment, autologous cell therapy, using CD133+ BMDSCs in conjunction with hormonal replacement therapy (HRT), increased the volume and duration of menses as well as the thickness and angiogenesis processes of the endometrium while decreasing intrauterine adhesion scores. Although the fertility outcome is still pending, this is a very important proof-of-principle study showing that this novel autologous cell therapy is a promising therapeutic option for infertility patients with these so far incurable endometrial pathologies. Its findings are in line with the recent reports by Jan Brosens' group on the role of endometrial stem cells in recurrent abortion. Gradually the endometrium is revealing (a few of) its secrets.
Mancy Tong, Torsten Kleffmann, Shantanu Pradhan, Caroline L. Johansson, Joana De Sousa, Peter R. Stone, Joanna L. James, Qi Chen and Larry W. Chamley, Proteomic characterization of macro-, micro- and nano-extracellular vesicles derived from the same first trimester placenta: relevance for feto-maternal communication, Hum. Reprod. (2016) 31 (4): 687-699. doi: 10.1093/humrep/dew004
The fetus communicates with the mother by shedding placental vesicles
How do fetus and mother communicate? Many theories exist, but their way of communicating is still for a large part mysterious. The first trimester human placenta sequesters extracellular vesicles of different size – small, tiny and minuscule – that are shed by the syncytiotrophoblastic apical plasma membrane into the maternal bloodstream. These vesicles circulate and participate in the fetomaternal cross-talk. Mancy Tong and her co-workers from Auckland, New Zealand, characterized the proteins within these shed syncytiotrophoblastic vesicles, classified according to their size: macro-, micro- and nanovesicles, all obtained simultaneously from the same first trimester placenta. Their study carefully dissected the proteome of each of the three categories of vesicles and identified some 1500 proteins; 1125 proteins were shared among all three types of vesicle, and up to 223 proteins were unique to each individual size category, suggesting that maybe trophoblastic macro-, micro- and nanovesicles are each likely to be produced by different mechanisms and have distinct effects on the maternal cells with which they interact. Each mediates communication between the fetus and its mother, a crucial process for both healthy and pathological pregnancies; protection or rejection of the fetus may result.
Not surprisingly Gene Ontology Pathway Analysis showed an enrichment of proteins involved in vesicle transport and inflammation in all three fractions of extracellular vesicles. In contrast, the expression levels of proteins involved in internalisation of vesicles (Annexin V, calreticulin, CD31, CD47), the complement pathway (C3, decay accelerating factor [DAF], monocyte chemoattractant protein [MCP], protectin) and minor histocompatibility antigens (ATP-dependent RNA helicase [DDX3], ribosomal protein S4 [RPS4]) were different between different sized vesicles. The authors propose that various extracellular vesicle fractions can interact with different maternal cells and have unique effects to mediate fetomaternal communication during early pregnancy. There is evidence that trophoblastic extracellular vesicles from normal pregnancies are anti-inflammatory and can induce maternal immune tolerance. Minor histocompatibility antigens on trophoblastic vesicles may be an important mechanism by which the maternal immune system is able to recognise and tolerate the fetal allograft during a healthy pregnancy, or alternatively, reject an abnormal pregnancy. It also may generate an undesirable destructive immune response in cases of recurrent miscarriage or recurrent placental abruption. Future functional studies are required to investigate the exact factors involved in these events, but this New Zealand study is an important step forward.
Cumulative live birth rates after three complete IVF cycles in the UK.
Something unique happened this month. We were preparing publication of the paper by McLernon, Maheshwari, Lee and Bhattacharya from Aberdeen on the cumulative IVF live birth rate in 178,898 women, as derived from the 1992-2007 HFEA data sets. They calculated a conservative estimate of a live birth after three completed cycles of IVF (including replacements of all resulting frozen embryos) of 42% for the treatments commencing in 1999-2007. The corresponding optimal estimate was 57%. Large numbers, solid figures, why not give them some extra exposure in an Editor’s Highlight?
Just when I was preparing this little piece of text, the December issue of the JAMA arrived. It carried an article by Smith, Tilling, Nelson and Lawlor from Bristol and Glasgow on the cumulative live birth rate in 156,947 women, as derived from the same HFEA data sets, but now from 2003-2010. They calculated a conservative estimate of live birth after three complete cycles of 45% and an optimal estimate of 59%.
What does this tell us? Three things: 1. After three cycles of IVF plus subsequent replacement of all resulting frozen embryos, at least 42% (conservative estimate) of couples have achieved a pregnancy; 2. There still occurs a slight increase of pregnancy rates over the recent years; 3. The HFEA produces stable figures, shares them dutifully, but does not hesitate to let two groups of investigators do the same work twice. If one group of authors had submitted the same two papers to two different journals one of them would have been labeled a “redundant publication”, a minor form of ethical misconduct. Not so now. Now it is a scientifically highly laudable – but accidental – validation of data.
George Griesinger, Beware of the 'implantation rate'! Why the outcome parameter 'implantation rate' should be abandoned from infertility research, Hum. Reprod. (2016) 31 (2): 249-251. doi: 10.1093/humrep/dev322
Abandon the Implantation Rate!
We publish a very provocative Opinion article by George Griesinger in Human Reproduction this month, a paper that in my opinion everyone should read. That’s why we made it ‘free access’ to all.
Griesinger argues in it that using the implantation rate (IR) indiscriminately in research reports is fraught with methodological pitfalls and interpretational difficulties. These are especially pertinent to studies in which systematically unequal numbers of embryos are transferred in groups being compared and where IRs are calculated on aggregate numbers. Embryo selection studies (e.g. by PGS, blastocyst culture, or by time-lapse morphology) are prime examples. Griesinger warns that more often than not the IR increases confusion rather than insight in physicians searching the clinical research literature for efficacious treatment options for their patients. He concludes that the IR should therefore not be used any longer in clinical trial reports. Read it!
Barbara Luke et al., Assisted reproductive technology use and outcomes among women with a history of cancer, Hum. Reprod. (2016) 31 (1): 183-189. doi: 10.1093/humrep/dev288
You may have followed the discussion in Human Reproduction on the increasing popularity of linking large administrative databases and drawing far-reaching conclusions. Often the linkage studies we receive at Human Reproduction seem like fishing expeditions, reporting increasingly improbable results: Obese women between 31 and 37, living in rural areas, having a fair skin, red hair and blue eyes, and using non-donor oocytes for IVF, have a higher than average pregnancy chance. I feel I also need to mention good things about such linkage studies, once an opportunity occurs. This month there is such an opportunity.
More and more women faced with a cancer diagnosis present for fertility preservation. In this month’s issue of Human Reproduction, Barbara Luke and co-workers, from East Lansing, Michigan, USA, address the issue how outcomes of women presenting for ART after cancer diagnosis compare to those of women without cancer. In a large database linkage study including 53,426 women they found that the likelihood of a live birth among women with prior cancer is reduced and varies by cancer diagnosis. It however is similar to women without cancer when donor oocytes are used. In women who only used their own oocytes, those with prior cancers were significantly less likely to have a live birth than those without cancer (adjusted odds ratio (AOR) 0.36 [0.28, 0.46]). This was also evident with specific cancer diagnoses: breast cancer (0.19 [0.11, 0.30]), cervical cancer (0.33 [0.13, 0.84]) and all female genital cancers (0.47 [0.25, 0.86]). The authors conclude that since the live birth rates using donor oocytes were not reduced in women with a prior cancer, but were reduced with their own oocytes, this suggests that factors acting in the pre- or periconceptional periods are responsible for the decline in live birth rates in patients with previous cancer diagnoses.