Fertility treatment and risk of cerebral palsy: has the association changed in Australia?

study questionWhat is the association between treated (ART, ovulation induction (OI)) and untreated subfertility and risk of cerebral palsy (CP)? SUMMARY ANSWER: There is no increased risk of CP in births to subfertile untreated women; the risk of CP has declined in ART conceptions (IVF, ICSI) and is greatest in births conceived using OI medications as a sole therapy. WHAT IS KNOWN ALREADY: Nordic data suggest a decline in risk of CP for ART births over time with the shift to single embryo transfer (SET), but there are no recent data from other countries and no Australian data for births after 2002. Few studies have examined the risk of CP in births conceived after OI or untreated subfertility; these comparison groups may assist in separating ART treatment effects from potential effects of subfertility itself. STUDY DESIGN, SIZE, DURATION: Retrospective cohort study using linked population-based data for women with a live birth (LB) in Western Australia from 2003 to 2014, corresponding to 10 126 ART births (3.1%), 4102 births after OI (1.2%), 11 554 births to subfertile untreated women (3.5%), and 305 508 (92.2%) births to fertile women conceiving naturally. PARTICIPANTS/MATERIALS, SETTING,

methodsStatutory ART, birth, hospital, CP, and pharmacy data were linked to identify four 'conception groups' (ART, OI, subfertile untreated, and fertile natural conceptions), outcome data, and potential confounders. CP descriptions were verified at 5 years of age. Birth prevalence of CP was calculated for each conception group per 1000 LB. Poisson regression with robust standard errors was used to estimate adjusted risk ratios (aRR) for CP with adjustment for maternal demographic characteristics, pre-existing conditions, and adverse obstetric history (with 95% CI). Births to fertile women were the reference group. Analyses included all children and singletons and twins separately. Weinberg's differential rule was used to estimate the rate of monozygotic twinning across our four conception groups and, for ART births, by length of embryo culture. MAIN RESULTS AND THE ROLE OF CHANCE: CP was diagnosed in 29 ART children (2.9/1000 LB; 95% CI 1.92-4.11), 16 children born after OI (3.9/1000 LB; 95% CI 2.23-6.33), 23 children in the subfertile untreated group (2.0/1000 LB; 95% CI 1.26-2.99), and 610 in the fertile group (2.0/1000 LB; 95% CI 1.84-2.16). When stratified by plurality and prematurity, risk was increased only for ART twins (aRR 2.8, 95% CI 1.42-5.39) born preterm (<37 weeks) and OI singletons (aRR 2.9, 95% CI 1.41-5.84) born preterm. Twins after SET comprised an increasing proportion of all ART twins over time, and of all ART twins with CP. Most twins after SET were same-sex twins (90.7%) and we estimated that 81% were monozygous, with the majority conceived after the transfer of a single blastocyst. The monozygotic twinning rate was higher after blastocyst-SET compared with cleavage-SET (prevalence ratio 1.7 (95% CI 1.07-2.73)). LIMITATIONS, REASONS FOR CAUTION: Small numbers of children with CP in subfertile groups prevented more detailed analyses of pregnancy complications and ART cycle characteristics. More recent data from larger populations are required to confirm whether risk of CP is increased following blastocyst transfer, in twins after SET, and following OI as a sole therapy. As zygosity data were not available, Weinberg's differential rule was used to estimate the rate of monozygotic twinning and may have led to some misclassification. WIDER IMPLICATIONS OF THE

findingsPrevalence of CP following ART has declined in Western Australia, as in the Nordic countries. Avoiding multiple embryo transfer remains important. Monozygotic twinning after SET may increase the risk of CP for ART twins and is more common after blastocyst compared with cleavage-stage transfer. Embryo culture, manipulation, and freezing strategies to reduce monozygotic twinning should be explored. Data about OI (as a sole therapy) are not routinely collected in most countries, but OI-conceived births had the highest risk of CP and may warrant closer scrutiny. Pre-treatment counselling should continue to promote SET while incorporating the small risk of embryo splitting; risk of multiple birth should be discussed with those planning to use OI. STUDY FUNDING/COMPETING INTEREST(S): This work was supported by the Australian National Health and Medical Research Council, grant number 1086530 (to M.H.) and The Research Foundation of Cerebral Palsy Alliance, grant number 06221 (to M.H.). The sponsors had no role in the design or conduct of the study, or the decision to submit for publication.Professor Hart is the Medical Director of Fertility Specialists of Western Australia and National Medical Director of City Fertility Clinic, a shareholder in CHA SMG; he has received accommodation support from Merck to attend ESHRE and has received travel and accommodation support, as well as educational sponsorship from MSD, Merck-Serono, Origio, Igenomix, and Ferring Pharmaceuticals. He has also received personal fees for presenting a Merck webinar. There are no other conflicts of interest to declare. TRIAL REGISTRATION NUMBER: N/A.