Change of HeART: Cardiovascular Implications of Assisted Reproductive Technology.

Although ART clearly is a safe and effective treatment, ART-conceived offspring may have a higher CV risk, compared to that of offspring conceived spontaneously. An important point to note, however, is that the existing literature on the association between ART and CV risk in offspring remains very limited and conflicting, making it difficult to formulate strong conclusions. A large European cohort study of over 7 million children followed for nearly a decade after birth reported no significant difference in CVD or type 2 diabetes mellitus between ART-conceived offspring and those conceived spontaneously. 132 However, ART-conceived offspring did have a significantly elevated risk for obesity. This finding was replicated in a prospective cohort study of 764 children aged 6-10 years, in which the ART-conceived group displayed an increased body mass index compared to that in the spontaneously conceived group. 133 Additionally, participants in the ART-conceived group demonstrated disadvantageous differences in left ventricular structure and function, including left ventricular hypertrophy. Similarly, other studies also have reported unfavourable changes to cardiac function in ART-conceived offspring. 134 , 135 Increased blood pressure and hypertension have been inconsistently described in the literature comparing ART-conceived offspring to spontaneously conceived offspring, 133 , 136 and a systematic review and meta-analysis of 19 studies concluded that ART-conceived offspring had a very small but statistically significant elevation in blood pressure, compared to that in offspring conceived spontaneously. 34 Finally, a systematic review and meta-analysis of 41 studies reported significantly elevated odds of congenital heart defects among ART-conceived offspring, as compared to the odds among those conceived spontaneously. 137 Adverse changes to vascular function also have been observed among ART-conceived offspring. In a cross-sectional study of 122 offspring in childhood, increased arterial stiffness and impaired endothelial function were observed in ART-conceived offspring compared to offspring spontaneously conceived. 138 In a follow-up study 5 years later, the vascular dysfunction observed at the initial assessment persisted, with the ART-conceived offspring exhibiting a persistent 25% lower flow-mediated dilation of the brachial artery, increased pulse-wave velocity, and increased carotid intima-media thickness. 136 Low birth weight is an independent risk factor for future CVD. 139 A number of systematic reviews and large cohort studies have addressed whether birth weight differs among offspring who were conceived spontaneously vs ART-conceived offspring. 140 , 141 , 142 , 143 , 144 , 145 , 146 Offspring conceived through ART tend to have lower birth weights than offspring conceived spontaneously, which may contribute to an increased future CV risk. This trend is persistent in both singleton and multiple pregnancies. 35 An intriguing finding is that this relationship between IVF-conceived pregnancies and low birth weight appears to be exclusive to fresh-embryo transfers, and no increased odds for low birth weight is demonstrated in IVF-conceived pregnancies with frozen-embryo transfer, according to a recent meta-analysis. 147

Infertility affects 16% of Canadians, and the use of ART is rising rapidly. 26 , 27 , 39 Nearly 30,000 ART treatments are completed in Canada each year, and approximately 2% of all live births in Canada are resultant from ART. 27 , 29 , 148 Despite the rapidly increasing prevalence of ART use, the CV implications of ART remain poorly understood. ART is associated with adverse CV outcomes along several points of the ART timeline ( Fig. 1 ), and this relationship is likely mediated by sex hormone changes, inflammatory changes, and changes to vascular health. Further, although it is clear that ART impacts short-term CV and pregnancy risk, whether ART is associated with adverse CV effects over the long-term remains unclear. Additionally, a full understanding of the CV health implications for ART-conceived offspring is lacking. Figure 1 Cardiovascular (CV) outcomes throughout the assisted reproductive technology (ART) timeline. COS, controlled ovarian stimulation; HDP, hypertensive disorders of pregnancy; OHSS, ovarian hyperstimulation syndrome; RAAS, renin-aldosterone angiotensin system. Cardiovascular (CV) outcomes throughout the assisted reproductive technology (ART) timeline. COS, controlled ovarian stimulation; HDP, hypertensive disorders of pregnancy; OHSS, ovarian hyperstimulation syndrome; RAAS, renin-aldosterone angiotensin system. Large-scale, well-designed prospective cohort studies are urgently needed to further evaluate the long-term CV implications of ART on both the individuals treated with ART and their offspring, as well as the contribution of differing ART protocols and ART success. Further, studies to identify the mechanistic pathways that link ART to CV and vascular outcomes are critically important to identify biomarkers that may predict ART-associated CV complications, and to test appropriate and personalized preventive strategies, including antthrombotic strategies, to limit each individual’s risk. Overall, although much important work has been completed to elucidate the relationship between ART and CV health, prioritization of further work is necessary in order to optimize CV risk reduction in females seeking ART care, as well as their offspring. Overall, ART has provided hope to millions of individuals with infertility worldwide. However, the relationship between ART, COS, and CV health is complex and multifaceted, necessitating further investigation to optimize the CV well-being of both females undergoing ART and their offspring. This future research holds the key to formulating evidence-based guidelines to mitigate potential CV risks for individuals treated with ART, and to providing critical knowledge and empowerment for the Canadian public.

ART refers specifically to fertility treatments that employ manipulation of oocytes or embryos, and this includes procedures such as in vitro fertilization and female fertility preservation. 28 ART procedures generally utilize COS, which most commonly involves administration of exogenous gonadotropins, including follicle-stimulating hormone (FSH), sometimes with the addition of luteinizing hormone, to stimulate supraphysiologic ovarian follicle growth. 58 Simultaneously, gonadotropin-releasing hormone analogs also are typically administered exogenously, to suppress the release of luteinizing hormone and subsequent ovulation, to allow for ongoing growth of follicles. 59 , 60 Gonadotropin and gonadotropin-releasing hormone analog administration continues until the follicles have reached a desired number and size, and then oocyte maturation is induced (usually through administration of human chorionic gonadotropin) prior to surgical retrieval of oocytes. 61 Estradiol is produced and secreted by ovarian follicles, and thus, COS also is marked by a substantial increase in endogenous estradiol levels, of up to 20 times baseline levels. 62 In ART, COS is followed by oocyte retrieval, in which transvaginal ultrasound-guided aspiration is employed to retrieve mature oocytes from the ovarian follicles. 63 Following retrieval, the oocytes may be either cryopreserved (fertility preservation) or fertilized with sperm in vitro (in vitro fertilization [IVF]). If the oocytes are fertilized, the resulting embryos may be transferred immediately to an individual’s uterus to initiate a pregnancy (usually at the cleavage stage [3 days following fertilization] or the blastocyst stage [5 days following fertilization]) or alternatively, cryopreserved for future transfer. 60 Multiple sex hormones that are critical to the COS process have been implicated directly in CV health and function. Endogenous estrogen has multiple complex effects on vascular physiology, including vasodilation and a reduction in blood pressure, mainly through activation of nitric oxide synthesis in endothelial cells and interaction with the renin-angiotensin-aldosterone system (RAAS). 64 , 65 , 66 Estrogen signaling also plays an important role in favourable lipid metabolism and dampening of inflammation, thereby providing protection from atherosclerotic CVD. 65 , 66 In fact, the rapid decline of endogenous estrogen is postulated to be a major factor in the increased CV risk exhibited by the population of postmenopausal females. 67 In contrast, the role of exogenous estrogen in CV health is not well understood. Oral formulations of exogenous estrogen, including hormonal contraception and hormone replacement therapy, have been associated with increased CV risk, although the effects of non-oral formulations are less clear. 15 , 68 The gonadotropin FSH also is independently associated with CVD in females. Although menopause is well understood to be strongly correlated with increased CV risk, which is most likely due in part to reduced estrogen levels, emerging evidence suggests that increased levels of FSH also may contribute independently to menopause-related CV risk. Higher FSH levels are associated with disadvantageous lipid profiles, 69 and independent of estrogen, higher FSH levels have been associated with an increased 10-year risk of atherosclerotic CVD. 25 FSH receptors have been found in vascular endothelium, suggesting that they play a direct role in vascular health, 70 , 71 and animal models demonstrate that FSH blockade is effective at reducing atherosclerotic CVD. 72 Ovarian hyperstimulation syndrome (OHSS) is a serious complication of COS and is characterized by ovarian enlargement, increased vascular permeability, and fluid shifts from the intravascular space to the third space, resulting in marked hemoconcentration. 73 , 74 , 75 The incidence of moderate-to-severe OHSS is approximately 1%-5% in all ART cycles, although it can be as high as 20% in higher-risk populations, such as those with PCOS. 76 With respect to its impact on the vascular system, OHSS results in severe endothelial dysfunction and increased vascular permeability. 74 Although the exact physiologic mechanisms have yet to be elucidated, OHSS appears to be mediated in part by vascular endothelial growth factor, expressed by granulosa and theca cells during COS. 77 , 78 , 79 , 80 Vascular endothelial growth factor appears to drive severe vascular permeability 77 , 78 and also may contribute to endothelial dysfunction. 81 The RAAS also plays an important role in OHSS pathophysiology. 75 , 82 Follicular fluid retrieved from female patients treated with ART complicated by OHSS reveals unusual levels of RAAS components, 83 and heightened systemic RAAS activity also has been demonstrated in OHSS, 84 , 85 although whether this effect is an independent mechanism or is simply related to third spacing and intravascular volume depletion is unclear. Interesting to note is that OHSS appears to be an exaggerated pathologic response to COS, and even COS cycles that are uncomplicated by OHSS are affected by a similar, albeit lesser, level of vascular dysfunction. 82 COS without OHSS has been associated with some level of vascular permeability, as well as heightened RAAS activity, and this degree of RAAS activation appears to correlate with the number of stimulated follicles. 86 , 87 , 88 This subclinical COS-associated vascular dysfunction may help to provide a mechanistic link to the heightened CV risk of COS and ART. Likely related to hemoconcentration, inflammation, hypercoagulability, as well as high levels of estrogen inherent to OHSS, arterial and venous thrombosis are common complications of OHSS. 89 In fact, females treated with ART complicated by OHSS have an up to 100 times increased risk of thrombosis, compared to pregnant individuals who conceived spontaneously. 90 Such thromboembolic events are often severe and may involve the cardiac and cerebrovascular arterial systems, resulting in myocardial infarction and stroke. 89 , 91 Although formal guidelines are lacking, consideration for prophylactic anticoagulation may be prudent in OHSS. 32 , 89 Even in the absence of OHSS, registry data suggest that individuals treated with ART have a heightened risk of venous thromboembolism, which often becomes apparent in the first trimester of pregnancy. 92 , 93 , 94 Specifically, a recent systematic review and meta-analysis of 14 studies reported that individuals treated with ART experienced a 2-3-fold increased risk of venous thromboembolism, compared to individuals who conceived spontaneously (relative risk [RR]: 2.66; 95% confidence interval [CI]: 1.60-4.43). 95 The elevated thromboembolic risk associated with ART is postulated to be related to the high-estrogen state of COS that directly precedes pregnancy in IVF with fresh embryo transfer. 96 This possibility is supported further by a recent study that demonstrated an 8-fold increased risk of venous thromboembolism in individuals treated with IVF who underwent a fresh embryo transfer (in which the embryo is transferred to the uterus immediately following COS), compared to the risk in individuals who conceived spontaneously. 97 This same risk differential, however, was not demonstrated in individuals treated with IVF who underwent a frozen embryo transfer (in which the embryo is transferred to the uterus following a period of time in which it is cryopreserved). The thromboembolic risk for individuals treated with unsuccessful ART cycles (cycles that did not result in pregnancy) remains unclear. Two studies have examined the relative thromboembolic risk of successful vs unsuccessful IVF cycles, with conflicting results. 98 , 99 A prospective cohort study of nearly 1000 women referred to a regional thrombosis unit reported a significantly lower risk of venous thromboembolism in individuals with unsuccessful IVF cycles. 98 Conversely, registry data from the Registry of Patients with Venous Thromboembolism (RIETE) suggest that although successful and unsuccessful IVF cycles were equivalent with respect to the incidence of venous thrombosis, individuals with unsuccessful cycles were at significantly increased risk for pulmonary embolism. 99 Although ART remains a generally safe procedure for pregnancy induction, many studies demonstrate that individuals treated with ART experience an increased risk of adverse pregnancy outcomes compared to individuals who conceived spontaneously. Hypertensive disorders of pregnancy (HDP) complicate up to 10% of all pregnancies on a global scale, 100 , 101 , and the risk of developing HDP, including preeclampsia, is increased in individuals who conceived through ART. 102 Multiple systematic reviews and meta-analyses have reported this finding, 103 , 104 , 105 , 106 and most recently, a 2021 meta-analysis of 85 studies demonstrated a higher odds of HDP for pregnancies conceived through ART, compared to those spontaneously conceived, including in singleton pregnancies (odds ratio [OR] 1.70; 95% CI 1.60-1.80) and in multiple pregnancies (OR 1.34; 95% CI 1.20-1.50). 107 The odds of HDP are reported to be even higher in ART treatments that utilize donor oocytes. 106 , 107 , 108 This HDP risk also may be escalated in individuals treated with multiple IVF cycles, those who have a higher number of embryos transferred, and those who have a longer duration of ovarian stimulation. 90 , 94 , 106 An interesting finding is that the risk of HDP also appears to differ between ART-conceived pregnancies with fresh-embryo transfer and those with frozen-embryo transfer. 109 Multiple meta-analyses that examined the odds of HDP occurring in ART also have assessed the relative odds in fresh- vs frozen-embryo transfer. 106 , 107 , 110 Consistently, ART-conceived pregnancies with frozen-embryo transfer have been reported to have increased odds of HDP compared to ART-conceived pregnancies with fresh-embryo transfer. The difference in pregnancy-associated outcomes between fresh- and frozen-embryo transfer has grown to be the subject of ongoing research, wherein the corpus luteum has been the focus. The corpus luteum is a temporary endocrine structure that forms after ovulation and produces progesterone and other factors to prepare the endometrial lining for embryo implantation and support early pregnancy. 83 One corpus luteum is present in pregnancies with spontaneous conception, and often more than one corpus lutea are present in ART-conceived pregnancies with fresh-embryo transfer. The corpus luteum is absent, however, in ART-conceived pregnancies with frozen-embryo transfer or donor oocytes, both of which demonstrate increased odds for HDP, and require exogenous progesterone administration for luteal endometrial support. 111 Data suggest that abnormal numbers of corpora lutea, such as in ART-conceived pregnancies, are associated with adverse vascular changes that may lead to adverse pregnancy outcomes, including HDP. Specifically, in ART-conceived pregnancies without a corpus luteum (ie, IVF with frozen-embryo transfer or donor-oocyte ART), the expected decline in arterial stiffness and pulse wave velocity is attenuated in the first trimester, and a reduction in endothelial function also is demonstrated. 112 , 113 Further, increased aortic augmentation also is demonstrated in the third trimester in ART-conceived pregnancies without a corpus luteum. 114 Notably, a relative increase in arterial stiffness throughout pregnancy also has been demonstrated in ART-conceived pregnancies with > 3 corpora lutea (ie, IVF with fresh-embryo transfer), compared to spontaneously conceived pregnancies. 112 , 114 These findings suggest that the corpus luteum and the factors it secretes, including relaxin, which is not detectible in pregnancies without a corpus luteum, may be critical contributors to the healthy vascular response to pregnancy. 83 , 115 The absence of relaxin, known to be important for endothelial function and vascular relaxation, may predispose individuals to increased arterial stiffness and HDP. 83 Further, in pregnancies with > 3 corpora lutea, increased RAAS activation has been demonstrated, indicating a potential role for the corpus luteum in healthy RAAS function. 83 , 116 In addition to HDP, other maternal pregnancy complications are present at an increased frequency in ART-conceived pregnancies compared to spontaneously conceived pregnancies. Specifically, ART-conceived pregnancies have a higher incidence of spontaneous abortion, 35 , 117 , 118 and maternal complications including gestational diabetes and abnormal placentation, known risk factors for future CVD, 100 , 119 , 120 occur at an increased frequency. 103 The long-term CV effects of ART remain unclear, and the available literature on this topic is limited. A large population-based cohort study of more than 240,000 Swedish women with a mean follow-up of 8.6 years reported a higher incidence of hypertension among individuals with pregnancies resultant from IVF, compared to those spontaneously conceived (HR: 1.27, 95% CI: 1.13-1.41), as well as a trend toward an increased incidence of stroke (HR: 1.27, 95% CI: 0.96-1.68). 121 Conversely, a large cohort of > 115,000 individuals in the Nurses Health Study reported no relationship between ART and hypertension later in life. 88 Further, a Canadian cohort study of over 1 million women demonstrated that in an adjusted analysis, women with pregnancies resultant from fertility therapy (ART and other types) had a lower HR of CV events compared to women with pregnancies related to spontaneous conception, and concluded that fertility therapy was not associated with an increased risk of CVD later in life. 122 Another interesting finding is that unsuccessful ART (ie, ART that did not result in a live birth) may predispose individuals to an increased CV risk, compared to successful ART. A population-based cohort analysis of more than 25,000 Canadian women treated with ART and followed for a median of 8.4 years, described a 21% relative increase in the annual CV event rate in women with unsuccessful ART, compared to the rate in women with successful ART. 123 This increased risk was driven primarily by heart failure (adjusted relative rate ratio 2.25; 95% CI 2.06-2.46) and stroke (adjusted relative rate ratio 1.25; 95% CI 1.15-1.37). Overall, the data remain inconclusive, and one meta-analysis has been completed to examine the risk of CVD in female patients treated with fertility therapy (ART and other types). 124 This meta-analysis reported no increased risk of a cardiac event (pooled HR: 0.91; 95% CI: 0.67-1.25) in women treated with fertility therapy; however, a trend toward a higher risk of stroke occurred (pooled HR: 1.25; 95% CI: 0.96-1.63). 124 All of the studies to date are limited by varying definitions of fertility therapy and ART, heterogeneity of comparator groups, variation in outcomes, and shorter follow-up periods. To complicate the picture further, females with infertility who seek treatment with ART may possess an inherently higher CV risk profile, which may contribute to the potential for increased CV risk observed following ART. Specifically, females with infertility, who are more likely treated with ART, may be older, may carry increased weight, and may have preexisting medical conditions, 42 , 125 all of which could contribute to an increased CV risk following ART. COS treatment and the associated hormone changes plausibly could result in long-term vascular effects in individuals treated with ART. During COS, female vascular function is modified, likely in direct response to changes in the hormonal milieu, though the long-term implications for these changes are not well understood. A meta-analysis of studies examining vascular function in response to COS reported a significantly increased heart rate and decreased blood pressure following COS. 126 Further, another study demonstrated significantly increased arterial stiffness during COS, as compared to baseline, independent of carotid artery diameter or blood pressure. 127 The effects of COS on endothelial function and integrity remain unclear. 128 , 129 Although changes to vascular function, including arterial stiffness and endothelial function, are predictive of long-term CV risk, 130 , 131 a point that remains unclear is whether the vascular health changes associated with COS are transitory or rather persist long-term. Further studies to address this question are urgently warranted.