JBRA Assist. Reprod. 2011;15(4):41-43
ARTIGO DE OPINIÃO
doi: 10.5935/1518-0557.2011.15.4.09
1Title of a presentation given at the IFFS Workshop, “Global impacts in infertility” on 21 May for REDLara in Rio de Janeiro, Brazil. The Workshop was supported by unrestricted educational grants to IFFS from Schering- Plough, Merck Serono, IBSA and CasMed and by REDLara.
ABSTRACT
Standard references may be classic papers, those based on large data sets or more simply and commonly, the best data set to the present. The latter may be superseded at any time, but the standard must be better. It is incumbent on all practitioners to study the literature and know about the best evidence available. This is continuing education. This paper provides a contemporary list of a number relevant to today’s issues with a brief description of each. They would all repay more detailed study as they set benchmarks for quality.
Keywords: standard references, natural fertility, Assisted Reproductive Technologies, ART
RESUMO
Padróes de referências podem ser publicações clássicas, aquelas baseadas em dados robustos ou, mais simples e frequentemente, os melhores dados disponíveis. Estes últimos podem ser substituidos a qualquer momento, mas o padrão deve ser sempre o melhor. Todos devem buscar a literatura especializada em busca da melhor evidência. Isto é educação continuada. Este artigo fornece uma lista contemporânea de um número de trabalhos relevante para os temas em estudo , com uma breve descrição de cada. Todos mereceriam um estudo mais detalhado, já que definem parâmetros fundamentais para a qualidade.
Palavras-chave: padrões de referència, fertilidade natural, reprodução assistida
INTRODUCTION
Thinking of “standard references” about natural fertility, papers such as that of Tietze (1957), of pregnancy rates unfettered in a Hutterite population, of Tietze (1968) after stopping intrauterine or oral contraception or of Vessey et al. (1976) in a population stopping barrier contraception, spring to mind. Although these are rarely quoted nowadays they remain benchmarks because of the unique nature of the population studied, the size of the study or the conceptual or organisational skill in carrying it out. Of more interest recently has been the pregnancy rate in untreated subfertile couples (Snick et al., 1997) in a rural environment and the rate of treatment seeking for infertility across the world as described by Boivin et al. (2009). Of course there is a place for these classical papers, as there is for a first report, the first randomised controlled trial of a treatment, the largest study with good statistics, a fine review or a metaanalysis. In a fast moving field, what should one keep to await a better paper, one that covers a more representative or better selected population or one that is large enough to support appropriate statistical analysis? It depends on the question asked. In some areas a single element cannot be isolated for long enough to build up the numbers to obtain an adequate result or there is too much variation in practice to make a result more widely applicable. So let me describe a number of papers that I think have current relevance and will continue to be influential. They all repay careful perusal by demonstrating good design and implementation and can act as models for subsequent studies. Perhaps this should be the interpretation of the “standard references” of the title. Allusions to them in this paper will be brief, but sufficient to indicate their importance.
Papers considered
Van der Steeg et al. (2008) collected data from 24 hospitals in the Netherlands and related Body Mass Index (BMI) to the time to natural pregnancy in 3029 subfertile patients using a smooth polynomial curve. They showed that a BMI of 21-29 was associated with a minimum time. Above 29 there was a 4% lower pregnancy rate achieved. Criteria for a normal semen analysis have been arbitrarily chosen, but the recent 5th Edition of the WHO Manual has also provided the 2.5, 97.5 and intermediate centiles for men whose partners conceived within 12 months of discontinuing contraceptive use as published by Cooper et al. (2010). These parameters exclude morphology, although limits for Kruger strict morphology are given separately. This is the standard international reference for andrology. The difficulty of showing statistical significance has been highlighted by Sripada et al. (2010) who show, using Receiver Operating Characteristic curves, that only motility and morphology are significantly associated with natural pregnancy. Another way of looking at fertility in the female is to measure the live birth rate using optimum semen quality as chosen in a donor insemination programme. De Brucker et al. (2009) showed cumulative rates for age groups from 20-45 over a 12 month period of treatment with particular emphasis on the 40-45 year age group.The efficacy of ovulation induction using gonadotrophins to achieve conception and live birth over 12 months was described by Balen et al. (1994) in those with cycle abnormalities, as shown by the clinical conditions of weight related amenorrhoea, polycystic ovaries and hypopituitary hypogonadism. They reached the pregnancy rates of natural conception as described by Tietze (1968).A later treatment format reported by Dovey et al. (2008), clomiphene (50-100mg) and intrauterine insemination (IUI), yielded cumulative pregnancy rates over 9 months in groups of < and > 35 years. Treatment was cancelled if a luteinising hormone (LH) surge was not detected using dipsticks, there was no follicle >15 mm or if a serum progesterone concentration >3ng/ml was detected, but the analysis was done by intention to treat. Goverde et al. (2000) had randomised patients between IUI using spontaneous ovulation, stimulated IUI and in vitro fertilisation (IVF) for unexplained infertility and male factor infertility and shown that there was no difference between any of the three treatments and that differences in the pregnancy rate per cycle over 6 cycles for each condition could be accounted for entirely by the age of the woman. Importantly this showed that IVF was three times the cost for the same outcome.Prediction of response has become important for identifying non-responding patients and those who have excess response, likely leading to ovarian hyperstimulation syndrome (OHSS). Nelson et al. (2007) demonstrated that serum Anti-Müllerian Hormone (AMH) could differentiate these conditions better than could serum Follicle Stimulating Hormone (FSH) and Jayaprakasan et al. (2010) using 3D vaginal ultrasound elegantly showed that identifying and measuring individual antral follicles correlated well with serum AMH. Those patients with polycystic ovaries had more readily distinguishable responses shown by serum AMH than by serum FSH or Antral Follicle Count (Nardo et al., 2009).It is becoming increasingly important to record the outcomes from treating infertility to influence the political decision to provide infertility treatment, including the various forms of assisted reproductive technologies (ART), as an integral part of a national health service. Data on outcome have been presented by Pinborg et al. (2009) following up patients from four large public hospitals in Denmark. That represented two thirds of all those achieving at least one delivery in the 5 years after treatment, which could include three reimbursed cycles of IVF. Outcome was recorded for each of natural conception and the results of treatment by IUI, IVF, intracytoplasmic sperm injection (ICSI) and frozen embryo replacement (FER); 69% achieved a delivery. Modelling the effects of IVF on a population to predict the number of first liveborn deliveries per 100,000 women trying for their first baby over 10 years, Habbema et al. (2009) concluded that offering IVF after one year’s infertility was less costeffective than offering it after 3 years’ infertility as the additional number of deliveries was so small.IVF takes many forms and the cumulative pregnancy and live birth rates following natural cycle IVF over 10 months have been reported by Nargund et al. (2001). Conventional stimulated cycles have been linked in the Massachusetts Society for ART (SART) registry and described at different ages using a patient’s own oocytes and using donor oocytes (Stern et al. 2010).National databases offer much larger numbers of treatment cycles and a full spectrum of conditions treated, so smooth out the peaks and troughs seen in smaller ones. The Australian database can be compared with the New Zealand, contrasting a reimbursed private system with a national health service system based on strict criteria for access. Farquhar et al. (2010) show the results of single embryo transfer (SET) and the overall results of cycles using a patient’s own oocytes, both by female age.By linking the ART and Birth Defects registries and in-patient records Hansen et al. (2008) showed that there was a fourfold risk of being admitted to hospital, (both neonatal intensive care and later) in the first year after an ART birth.Although some countries now predominantly use SET, many do not. Although a meta-analysis emphasises the dilemma by showing that although a double embryo transfer (DET) slightly increases the live birth rate it markedly increases the rate of multiple pregnancy (Gelbaya et al., 2010). On the other hand Veleva et al. (2009) in an equally large data set clearly show that when frozen embryo transfer (FER) is also available elective SET markedly increases the cumulative live birth rate and dramatically reduces the multiple birth rate per oocyte retrieval. Subsequent frozen/thawed embryo replacement does not increase the total number of fresh and frozen transfer cycles per woman. As the cost of a FER is about 20% of that of IVF/ICSI, it reduces the overall cost and achieves a slightly better overall live birth rate than does DET. Evaluation of perinatal outcome after FER indicates that there is no increase in perinatal or infant mortality; the babies are more likely to be larger and less likely to be smaller than after fresh embryo transfer (Pelkonen et al., 2010; Pinborg et al., 2010). Although the reason for having oocyte donation is because a patient’s own oocytes are unlikely to result in a pregnancy, increasing age of the recipient of frozen embryos resulting from oocyte donation also results in reduced pregnancy rates (Dessolle et al., 2009).Attempted culture to blastocyst stage does not always result in having an embryo to transfer. However, using the Australian national database a comparison of outcome of cleaved embryo transfer at an earlier stage and blastocyst transfer at a later stage shows that selection of a single blastocyst results in a significantly greater delivery rate of a healthy baby (>37 weeks, ≥2.5Kg, surviving at least 28 days and without congenital abnormality) than any form of DET (Wang et al., 2010).Examining the total costs of an ART service within a national programme (Griffiths et al., 2010) demonstrates clearly that the overall cost of having a first baby at aged 42-45 is at least fourfold the cost of having one at 30-33 years of age. Similarly the older age group costs at least sixfold that of the younger for a second baby. Intermediate ages reflect intermediate costs, which in some systems have an effect on decisions about access.
CONCLUSION
These few examples of high quality papers across the spectrum of ART underline the diversity of material and the fact that almost all are recent. “Standard” papers referenced should be the best that are available and reflect a snapshot in time. The list requires selection from an increasing number of published papers and puts the onus on practitioners to keep up to date by reading at least the main relevant journals. This is continuing education, a demanding but enjoyable task.