AC TA Obstetricia et Gynecologica

AOGS COMMENTARY Is recurrent a useful clinical concept? WILLEM VLAANDEREN Malden, the Netherlands

Key words Abstract Miscarriage, , , genetics, chromosomal anomalies, recurrent Many treatments have been used to increase the chances of an ongoing preg- miscarriage nancy after recurrent miscarriage (RM). Yet no clear evidence for an effective intervention has been found. Therefore, the clinical concept RM should be Correspondence reconsidered. Both the statistical and clinical bases for RM are poor. If the Willem Vlaanderen, MD, Gynecologist individual risk is taken into account, then simple calculations explain the (retired), Vinkenlaan 24, 6581 CK Malden, observed frequency of RM. Most , either sporadic or recurrent, are the Netherlands. E-mail: [email protected] abnormal in construction or development. Although some risk factors for mis- carriage have been identified, such as age and number of previous miscarriages, Conflict of interest the importance of other factors remains uncertain. RM has a favorable progno- The author has stated explicitly that there are sis: the cumulative live birth rate is usually over 90% after two or three more no conflicts of interest in connection with this attempts. So maternal systemic pathology does not seem to play a major role article. in the cause of RM. Little benefit can be expected from any intervention aimed at increasing the chance of a live birth. RM does not appear to be a real patho- Please cite this article as: Vlaanderen W. Is recurrent miscarriage a useful clinical logical entity that requires a special explanation. Moral support should replace concept? Acta Obstet Gynecol Scand 2014; medical interventions in most cases. 93: 848–851. Abbreviation: RM, recurrent miscarriage.

Received: 26 November 2013 Accepted: 14 July 2014

DOI: 10.1111/aogs.12453

Introduction from sporadic miscarriage, or is it merely an example of “clustering illusion”? Recurrent miscarriage (RM, habitual abortion in older literature) is defined as three consecutive miscarriages Statistical aspects before 20 weeks of pregnancy (1). It comes as a major disappointment for the couples involved and most of Early miscarriages occur in approximately 15% of pregnan- them ask for an explanation and, if possible, treatment. cies. Therefore the risk of three consecutive miscarriages Over the years many causes for the condition have been has been calculated as 0.15 9 0.15 9 0.15 = 0.34%. The suggested, such as infections, vitamin deficiencies, endo- observed incidence of RM is approximately 1% of all fertile crinological and immunological disorders, psychological couples (1). When these figures are compared it is often factors, and most recently thrombotic events. Treatments concluded that RM cannot be merely a matter of chance based on these presumed causes have usually resulted in a (3,4). However, this conclusion is flawed for two reasons. live birth rate of 70–80%, an apparent improvement First, the individual miscarriage risk is not the same for compared with the patient’s history. However, controlled everybody and for every pregnancy. The risk is found to trials have failed to demonstrate a benefit of these inter- increase with each subsequent miscarriage. The recurrence ventions (2). These negative results should lead to recon- rate after three miscarriages is reported to be 25–30% sideration of the concept of recurrent miscarriage. Is RM (5–7). The recurrence rate after one or two miscarriages a special condition that requires a different explanation will be somewhere between 15 and 25–30%, cautiously

848 ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 848–851 W. Vlaanderen Recurrent miscarriage estimated as 18% after one and 21% after two miscarriages. of aneuploidy. A very small group, carriers of balanced Therefore the risk of three successive miscarriages is some- chromosomal translocations, was found to have a risk of thing like 0.15 9 0.18 9 0.21 = 0.57%. Secondly, the 40% in subsequent (14). The importance of denominator in the fraction of the 1% for RM is not other possible risk factors is uncertain. An association has the total number of women with three pregnancies, but the been found between RM and “anti-phospholipid syn- number of all fertile couples. Therefore, “secondary” RM drome”, but the causality is questionable, the working (child-m-m-m) contributes to the incidence as well. In the mechanism unclear, and the treatment results disputed 85% of women who start with a live birth, another (2,15,16). Anatomical variations such as a subseptate 0.85 9 0.57 = 0.48% of RM cases could be expected. So are suspected of increasing the risk, but positive the expected incidence of RM will be approximately treatment results are still missing. Uncertainty exists 0.57 + 0.48 = 1.05%. This is comparable to the figure about the role of overweight, smoking, thrombophilia observed, confirming that if the individual risk is taken into and thyroid disease. account, RM can be explained as a chance event. Overall, little clinical evidence exists to support the definition of a special pathological condition after three Clinical aspects miscarriages. Some investigators propose redefining RM to include just two miscarriages and not necessarily Early miscarriage usually seems to be part of the selection consecutive ones. However, the small and gradually process in human reproduction. Many conceptions disap- emerging differences between sporadic miscarriages and pear, both before and just after implantation. Analysis of recurrent miscarriages make a distinct cut-off point early miscarriages shows that most of these are defective impossible. No special risk factors have been identified in conception products. Chromosomal aberrations have the majority of women with more than one miscarriage. been found in 60–80%, with most of the remaining cases These miscarriages are therefore termed “idiopathic” or being anembryonic amniotic sacs or with serious “unexplained”. The use of these terms is confusing: there malformations. Fetuses that are both cytogenetically and is no valid reason to call a third miscarriage more unex- morphologically normal are found in at most 7% of all plained or idiopathic than the first miscarriage. The cen- miscarriages (8,9). tral question is why some women seem to have a larger Repeated miscarriages do not appear to be much different risk for a miscarriage than others. Identifying risk factors from the sporadic ones. Chromosomal anomalies have been might be useful in some cases, but denominating such found in 51% of RM cases compared with 76% in sporadic factors as a cause or an explanation for RM is an over- miscarriages (10). Therefore some investigators suppose simplification. that maternal causes could be more frequent than in spo- radic miscarriage. But the majority of RM are cases of Prognosis embryonic loss, meaning pregnancies where no cardiac activity has been identified (5,11). This suggests that intrin- The strongest argument against the existence of a disease- sic embryologic abnormalities are more likely than maternal like entity called RM is the favorable prognosis in most causes. There is a paucity of data regarding the morphology women with RM, as was found for instance by Clifford of the pregnancy product in recurrent miscarriages. No et al. (7). In a group of 129 women under 40 years with studies could be found that reported a larger proportion of a history of fewer than six “unexplained” miscarriages, potentially viable fetuses in RM than the 7% cited above. just 21% miscarriages occurred in the next pregnancy, Some investigators have suggested that RM is not being without medical intervention other than supportive care. caused by an increased maternal tendency to lose normal The success rate of subsequent pregnancies in 222 , but rather the opposite: too easy an acceptance of patients with “unexplained” RM in relation to both the compromised embryos during the implantation (12,13). number of previous miscarriages and age was studied by Some risk factors of clinical importance have been Brigham et al. (6). Among 30-year-old women, the next identified. The most outstanding feature is that the recur- pregnancy after three miscarriages was successful in 80% rence risk increases with the number of miscarriages of cases, after four miscarriages in 76% of cases and after in a woman’s history (5–7,10). The reason is most probably five miscarriages in 71% of cases. Among 40-year-old that some women have a lower and others a higher risk women, those percentages were 64, 58 and 52%, respec- than average. As women continue their attempts to achieve tively. Using these figures, the cumulative success rate in an ongoing pregnancy, the low-risk women disappear faster women deciding to continue their attempts after several from the RM population than the high-risk ones. miscarriages can be predicted. In a group of 100 women The risk of recurrence also increases with maternal age aged 30 years with three previous miscarriages, approxi- (5–7,10), which is largely explained by the increasing risk mately 80 women will have a successful subsequent

ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 848–851 849 Recurrent miscarriage W. Vlaanderen pregnancy and 20 will have a fourth miscarriage. If those continue their attempts. The best advice for most if not 20 try again, 76%, or 15, will experience an ongoing preg- all patients, is to accept the natural course of their preg- nancy. Of the five remaining women, 71%, or three to nancies: it has no side-effects, it does not interfere with four, will be successful in their sixth pregnancy. So the the selection process, and it is highly effective as long as expected cumulative success rate is 80 + 15 = 95% in two fertility is normal. Women with such an unlucky experi- pregnancies following the first three miscarriages, and 98– ence deserve all kinds of moral and practical support. 99% in three subsequent pregnancies. The same calcula- The most important part of that support is explanation tion for the group of 40-year-old women results in a of the basic facts: almost all miscarriages are defective, no cumulative success rate of 85% after two and 93% after proven treatments are available, the prognosis without three pregnancies. treatment is excellent. From these calculations it will be clear that if a woman Our knowledge of miscarriages still has many gaps. We has the courage to continue her attempts, the chance for would like to have clear answers to questions like: why success is very high. Even if the miscarriage risk is assumed are so many conceptions abnormal, why are some com- to be as high as 50%, the cumulative success rate in the promised embryos able to implant and develop for some subsequent pregnancies will be 50 + 25 + 12.5 = 87.5% in time, and why does this occur more often in some three, and up to 94% in four pregnancies after the first women than in others? Simply coming up with a new three miscarriages. In my own study of a group of unse- “explanation” and then testing this in randomized con- lected patients with three consecutive miscarriages, a trolled trials using groups of unselected RM patients does cumulative success rate of 96% was found (23 of 24) (17). not lead to conclusive results. A better understanding of Four of them had a total of five miscarriages before they the pre-conceptional part of the reproduction process were able to start a family. might ultimately allow us to find ways of reducing the This high success rate in subsequent pregnancies indi- miscarriage risk in individual cases. cates that most women with RM do not have systemic pathology that interferes with the development of a nor- Funding mal pregnancy. Incidentally, a miscarriage could be caused by maternal factors, but even then the chance for No specific funding. success in one of the subsequent pregnancies is high. The cumulative success rate was the same in both transloca- References tion carriers and non-carriers, namely 84% (12). If three consecutive miscarriages are taken as the cut- 1. Jauniaux E, Farquharson RG, Christiansen OB, Exalto N. off point for defining recurrent miscarriage, then the Evidence based guidelines for the investigation and group of women included will still contain too many medical treatment of recurrent miscarriage. Hum Reprod. low-risk patients. Including women with two miscarriages 2006;21:2216–22. might have an even more diluting effect. Without a stric- 2. Duckitt K, Qureshi A. Recurrent miscarriage. Clin Evid. – ter selection, any clinical trial in such a group will almost (online) 2011;1409 14. certainly yield a negative result. If amenable maternal 3. Saravelos SH, Li TC. Unexplained recurrent miscarriage: – causes seem to exist, then the disadvantages of a possible how can we explain it? Hum Reprod. 2012;27:1882 6. intervention should be weighed against the calculated 4. Larsen EC, Christiansen OB, Kolte AM, Macklon N. New prognosis. insights into mechanisms behind miscarriage. BMC Med. 2013;11:54. 5. Bricker L, Farquharson RG. Types of pregnancy loss in Conclusion recurrent miscarriage: implications for research and clinical practice. Hum Reprod. 2002;17:1345–50. RM does not seem to be a useful clinical concept. Some 6. Brigham SA, Conlon C, Farquharson R. A longitudinal women have more chance of a miscarriage than others. study of pregnancy outcome following idiopathic recurrent The observed recurrence rate is fully explained by chance miscarriage. Hum Reprod. 1999;14:2868–71. if the individual risk is taken into account. The majority 7. Clifford K, Rai R, Regan L. Future pregnancy outcome in of repeated miscarriages are just as defective as the spo- unexplained recurrent first trimester miscarriage. Hum radic ones. Women with RM have an excellent prognosis Reprod. 1997;12:387–9. for a live birth. Screening for factors that possibly 8. Philipp T, Philipp K, Reiner A, Beer F, Kalousek DK. increase the risk does not appear to be very useful, except Embryoscopic and cytogenetic analysis of 233 missed in individual cases, such as repeated loss of apparently : factors involved in the pathogenesis of normal fetuses. With or without extra risk factors, most developmental defects of early failed pregnancies. Hum women will be successful if they can be convinced to Reprod. 2003;18:1724–32.

850 ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 848–851 W. Vlaanderen Recurrent miscarriage

9. Ljunger E, Stavreus-Evers A, Cnattingeus S, Ekbom A, -maternal interactions and causes recurrent Lundin C, Anneren G, et al. Ultrasonographic findings in pregnancy loss. PLoS ONE. 2010;5:e10287. spontaneous miscarriage: relation to euploidy and 14. Franssen MTM, Korevaar JC, van der Veen F, Leschot NJ, aneuploidy. Fertil Steril. 2011;95:221–4. Bossuyt PMM, Goddijn M. Reproductive outcome after 10. Osagawara M, Aoki K, Okada S, Suzumori K. Embryonic chromosome analysis in couples with two or more karyotype of abortuses in relation to the number of miscarriages: case-control study. BMJ. 2006;332:759–63. previous miscarriage. Fertil Steril. 2000;73:300–4. 15. Laskin CA, Spitzer KA, Clark CA, Crowther MR, Ginsberg 11. Morikawa M, Yamada H, Kato EH, Shimada S, Yamada T, JS, Hawker GA, et al. Low molecular weight heparin and Minakami H. Embryo loss pattern is predominant in aspirin for recurrent pregnancy loss: results from the miscarriages with normal chromosome karyotype among randomized controlled HEPASA trial. J Rheumat. women with repeated miscarriage. Hum Reprod. 2009;36:279–87. 2004;19:2644–7. 16. Clark CA, Laskin CA, Spitzer KA. Anticardiolipin 12. Quenby S, Vince G, Farquharson R, Aplin J. Recurrent antibodies and recurrent early pregnancy loss: a century miscarriage: a defect in nature’s quality control? Human of equivocal evidence. Hum Reprod Update. Reprod. 2002;17:1959–63. 2012;18:474–84. 13. Salker M, Teklenburg G, Molokhia M, Lavery S, Trew G, 17. Vlaanderen W, Treffers PE. Prognosis of subsequent Aojanepong T, et al. Natural selection of human embryos: pregnancies after recurrent spontaneous abortion in first impaired decidualization of endometrium disables trimester. BMJ. 1987;295:92–3.

ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 848–851 851