ICSI is one of the commonest techniques practiced in ART laboratories all over the world. Its application on human gametes was first reported in 1988 [4]. The main indication for performing ICSI is male factor infertility; however, there is a trend of more widespread use of this technique even for patients not presenting with male factor infertility. In 2000, 47.6 % of ART was reported using ICSI, in 2006 the proportion increased to 66 % [5]; the level of this proportion has remained the same up to now [5, 6]. For those without male factor infertility in the US, the proportion of its use has increased from 15.4 % in 1996 to 66.9 % in 2012 [6].

There is a concern that ICSI may have detrimental effect on pregnancies and children. ICSI has removed the natural selection of the fertilizing sperm, and has also allowed the transfer of gene that would not normally be passed on. The technique may also inflict physical damage to the gametes [7]. A study on constitutional DNA copy number has detected a higher rate of changes in ICSI children in comparison to naturally conceived children; however, it is difficult to ascertain whether these genetic changes have any significant phenotypic consequences to the offsprings [8]. By comparing to conventional IVF, it has been reported that ICSI does not increase the risk of major birth defects [7, 9, 10]. In comparison to naturally conceived pregnancies, ICSI pregnancies have similar mean gestational age at birth, birth weight, neonatal distress level and Neonatal Intensive Care Unit (NICU) admission [11, 12]. A few studies have looked into the longer-term development of ICSI children. The studies of cognitive abilities, socio-emotional development and motor skills scores have been reported for children up to the age of 10. It is reassuring to learn that ICSI children largely performing on par with naturally conceived children [11–14]. In one study, IVF children were detected to have better simultaneous mental processing ability compared to ICSI children [13]. Interestingly, one of the studies showed that ICSI children had better interactional ability and lower distress level than naturally conceived children [15]. A few researchers studied the health, growth, and also pubertal and endocrinological changes on ICSI children up to late puberty. Basatemur et al. studied the growth of ART children up to the age of 18 years old. They compared 143 IVF and 166 ICSI children with 173 matched naturally conceived controls and concluded that there were no significant differences in head circumference, height and weight between the groups [16]. Pubertal development by Tanner stage and age of menarche have also been studied in ICSI children by Belva et al. Development of these sexual characteristics is largely similar for singleton born ICSI boys and girls in comparison to their 14-year-old spontaneously conceived counterparts; only one difference was detected: ICSI females had less pronounced breast development by comparison [11, 17]. In the same study, the authors also reported increased central, peripheral and total adiposity in ICSI children comparing to spontaneously conceived children. In advanced pubertal stages ICSI adolescents had more peripheral adiposity. The same group has also performed other studies to assess different aspects of health effect of ICSI on children. In one paper, they compared blood pressure between ICSI boys and girls with spontaneously conceived children before and after subjecting the participants to a stress test; no detrimental blood pressure effect was detected for ICSI children in the study [18]. To complement the studies on physical characteristics, the group measured salivary cortisol in the children [19]. They related alterations in cortisol level with changes in adiposity, blood pressure and glucose tolerance. In that study of 14-year-olds, ICSI females but not males were detected to have lower salivary cortisol concentrations in comparison to spontaneously conceived children. Belva et al. have also studied salivary testosterone concentrations in pubertal ICSI boys compared with spontaneously conceived boys. As ICSI removes the process of natural selection of sperms for fertilization, some worry that this technique may increase the risk of male offsprings inheriting karyotypic anomalies or Y-chromosome microdeletions from their fathers, resulting in genital malformations or impaired testicular function. The authors used salivary testosterone level as a surrogate marker of testicular function in 14 year-old male adolescents. In this study 58 ICSI male teenagers were compared to 62 spontaneously conceived counterparts. They found that the testosterone levels for these ICSI boys conceived from men with severely compromised spermatogenesis were similar to the naturally conceived group. In other studies, no increase in male urogenital anomalies was reported for ICSI children in comparison to the background risk [20, 21]. For children who inherit Y-chromosome deletions, it is suggested that the size of the deletion is not increased in the offspring, and the extent of infertility for the children is likely to be the same as for the fathers [20, 22].

In summary, the findings of the studies regarding ICSI pregnancies and children are generally reassuring. In comparison to spontaneously conceived offspring, ICSI adolescents are at risk of developing obesity [17]. However, from this study it is difficult to ascertain whether the phenotypic pattern is due to the effect of ICSI; previous study has shown that IVF children have more peripheral adiposity in comparison to spontaneously conceived children [23]. For the application of ICSI, various sperm selection techniques have been developed, most of them still lacking evidence in proving their safety and effectiveness [24–26].

The application of ICSI technique has permitted the use of surgically retrieved sperm for ART treatment. It is now possible for men with obstructive azoospermia, non-obstructive azoospermia and severe oligospermia to father children using non-ejaculated sperm. Successful testicular sperm retrieval has been reported in males with Klinefelter’s syndrome, and also in patients post chemotherapy [27–29]. The use of surgical retrieved sperm is not without concern. Some believe that ICSI using surgically retrieved sperm is a step further in eliminating the natural selection of suitable sperms for fertilization. In addition, sex chromosome anomalies and Y-microdeletions have been detected in more than 10 % of patients with non-obstructive azoospermia and oligospermia [30]. In testicular sperm extraction (TESE), immature testicular sperms can be extracted and used for ICSI. On the other hand, aged epididymal sperm from epididymal sperm aspiration (MESA) may contain chromosomal error [20]. It is feared that the genetic or chromosomal anomalies can be passed on to the offspring; ICSI using suboptimal sperm in theory can also have adverse effects on the pregnancies or children. Several studies reported the pregnancy outcomes of pregnancies from surgical retrieved sperm, by comparing them to ICSI pregnancies using ejaculated sperm, and also IVF and naturally conceived pregnancies. The results are largely reassuring. No significant differences have been reported in the rates of miscarriage, ectopic pregnancy, intrauterine growth restriction, maternal complications, preterm delivery, low birth weight, neonatal unit admission, perinatal mortality and infant mortality [7, 31–35]. Fedder et al. reported a lower caesarean section rate for the group of pregnancies from surgically retrieved sperm in comparison to IVF and ICSI using ejaculated sperms. [33] There is one finding of an increase in perinatal death for twins from surgically retrieved sperm when compared to ICSI using ejaculated sperm. [35] Some authors made comparison of the complications between obstructive and non-obstructive azoospermia in surgically retrieved sperms. The gestational age at birth, birth weight and neonatal outcomes are similar between these two groups [36]. There is a non-significant increase in miscarriage in the non-obstructive azoospermic group [37]. The miscarriage rate was the same regardless of whether the surgically retrieved sperm were from the testicles or epididymes [37]. In relation to fetal malformations, the results are more conflicting. The difference in results may be due to the variation in the definitions and categorizations used for different studies. Most studies reported no difference in the rate of congenital malformations [20, 34–36]. Guo et al. discovered a non-significant increase in birth defects in the TESA group (103 children) in comparison to 1008 children born after ICSI with ejaculated sperms. [32] Fedder et al. studied different groups consisting of 466 children born with surgically retrieved sperms, 8967 ICSI children with ejaculated sperms, 17,592 IVF children and 63,854 naturally conceived children [33]. By tests of variance, they reported the rate of undescended testicles and cardiac malformations in boys significantly increased from natural conception to IVF to ICSI with ejaculated sperm to ICSI with surgically retrieved sperm. In a different study, this research team also discovered an increased rate of hypospadias for children born with surgically retrieved sperm (three out of 197) [38]. However, in this study no direct comparison was made for this group of patients with children born with ICSI using ejaculated sperm.

Limited data are available on the genetic and chromosomal abnormalities for children born with surgically retrieved sperm. One study reported no difference in anomalies in pre- and post-natal karyotypes in viable ICSI pregnancies between surgically retrieved and ejaculated sperm. [35] Even for males with Klinefelter syndrome, 59 % of the embryos fertilized with testicular sperms were confirmed to have a normal karyotype [27]. Another study reported 100 % normal karyotype for 16 babies born using testicular sperms from males with Klinefelter syndrome [28]. At present there is no concern regarding the neurodevelopment of children conceived with surgically retrieved sperm. These children perform very well in the assessment of their milestones and skills [7, 34, 39]. A large number of surgically retrieved sperm samples were frozen and thawed at a later date for ICSI. It has been reported that the freezing of testicular sperm does not have any adverse effect on neonatal outcomes [28, 40].

Parents planning to undergo surgical sperm retrieval must be provided appropriate counseling. Most pregnancies resulted from this technique are uncomplicated, and the babies born are healthy. However, there are reports of an increased risk of congenital malformations; the risk can be as high as 8–10 % [20]. Males detected to have Y-chromosome deletions may also pass on the genetic malformation to the male offspring. Consideration should be made to establish a system to educate the family, and perhaps the children at a suitable age regarding the possibility of inherited infertility.

Phenotypic or genotypic changes may be induced in embryos subjected to various environments in vitro. Questions have been asked about whether the culture condition, media used, or length of culture have any significant impact on pregnancies and offspring. Some authors have shown that different culture media used can affect birth weight; however, not all studies demonstrated the same effect [41, 42].

In vitro maturation (IVM) is a technique becoming more popular in ART laboratory. There is a strong argument for its use in specific populations, for example in high responders to prevent ovarian hyperstimulation syndrome. Due to the lack of RCT, the safety of this new technique introduced in the early 1990s is in doubt. Studies with small samples did not detect any increase in the risks of fetal malformations or adverse perinatal outcomes [43, 44]. In a laboratory study, Yoshida et al showed that at cleavage stage the metabolic state (oxygen consumption) of embryos resulted from IVM and controlled ovarian hyperstimulation (COH) were the same [45]. In IVM babies, no abnormality was found in the expression of imprinting genes [45]. In a different laboratorial study by Virant-Klun et al., the authors detected some changes in the gene expression profile of oocytes in IVM [46]. It is unknown whether the shifts in the gene expression profile have any effect on pregnancy outcomes and the health of offspring.

Many ART centers culture the embryos today 5/6 for blastocyst transfer, with the aim to increase the success rates. There is some evidence that blastocyst transfer results in less miscarriage [47]. However, the practice of extended culture of embryos is not without health risks. It has been shown that this technique increases the likelihood of preterm delivery with odds ratio of up to 1.32 [48, 49]. Several studies also demonstrated that the transfer of blastocysts increases the chance of monozygotic twinning [50–52].

One of the most popular new technologies in ART is the time-lapse incubator. It is claimed that chromosomal normal and abnormal embryos have different morphokinesis (kinetic behavior), and the use of time-lapse technology can differentiate these embryos and its application can improve success rates [53, 54]. A decrease in miscarriage rate has been reported with the use of this technology [55]. In theory, the use of this technology should not pose any harm as its application is not truly invasive to the embryos in culture. However, prospective studies are underway currently and the data should be studied carefully when they become available [56].

In short, to date the available data have shown that the technique and duration of embryo culture can influence pregnancy outcome. It is undeniable that there is a lack of robust studies, and the evidence to support the safety of the new techniques in ART laboratories is missing.

Different embryo micromanipulation techniques have been introduced in ART laboratories. One of the techniques is assisted hatching; the use of assisted hatching is increasingly common [57]. There is good evidence that assisted hatching improves clinical pregnancy rate in poor prognosis patients, including those with prior failed IVF cycles [58]. It is debatable whether this technique increases miscarriage rate [57, 59]. As this process involves the disruption of the zona pellucida, a few studies showed an increase risk of dichorionic monozygotic twinning, especially if assisted hatching is performed on day 2–3 embryos [50–52, 60]. In a study to assess the safety of assisted hatching, Zhou et al. looked at 392 infants in total. The authors concluded that it did not make any significant difference in mean gestational age, mean birth weight and mean Apgar score for either singleton or multiple gestations [61].

In preimplantation genetic screening/diagnosis (PGS/PGD), cells or polar bodies are biopsied from the embryos; logically the removal of one or more blastomeres may adversely affect the development of an embryo. However, embryonic cells are totipotential in nature and perhaps the other remaining cells in the embryo have the capacity to accomplish different developmental pathways for the embryo to grow normally [62]. The safety data of this embryo manipulative technique have been reported in a few studies. In comparison to ICSI pregnancies, the rates of intrauterine growth restriction, low birth weight, congenital malformations, neonatal hospitalization, neonatal intensive care admission and perinatal death for singletons are similar [62–66]. Fewer multiple pregnancies following PGD presented with low birth weight (<2500 g) [64]. On the other hand, there is a report of an increase in perinatal deaths in post PGS/PGD multiple pregnancies [66]. In several studies, PGS/PGD has been shown to reduce the rate of miscarriage for patients with recurrent miscarriage; this includes parents with reciprocal or Robertsonian translocation [67–70]. Research on the neurodevelopment of children following PGS/PGD has produced some interesting results. Schendelaar et al. studied 49 children born following ART with PGS, comparing them with 64 children born following ART without PGS. The authors concluded that there was no difference in the neurodevelopmental outcome of these children [71]. In a different study, a Dutch group reported that PGS is not associated with any changes in mental, psychomotor and behavioral outcomes at 2 years in children born after PGS. Scores on all tests were within normal range. However, when compared to children born after IVF without PGS, PGS children had lower neurologic optimality scores, this may be a signal of less favorable long-term neurologic outcomes in these children [72]. In a separate paper, the same group reported similar neurologic outcome before 18 months for ART children with or without PGS. At 18 months, they reported increased frequencies of dysfunction in fine motor abilities and posture, and also muscle tone dysregulation in PGS children [65].

Embryo manipulation techniques, specifically assisted hatching and PGS/PGD, may be useful in improving the ART success rates for specific groups of patients. However, one should be aware of the potential adverse effects. Reassuringly both of these techniques have not been shown to cause significant problems in pregnancy and perinatal period for the majority of patients. However, one should be cautious in applying these techniques unselectively. Assisted hatching, especially performed on cleavage stage, increases the chance of monozygotic twinning. An increase in perinatal deaths was also reported in PGS/PGD multiple pregnancies. In addition, there is also a concern regarding the long-term neurologic outcomes of children following PGS/PGD.

The technique of oocyte cryopreservation has undergone tremendous improvements in recent years. For many years, the practice of this technique is lagging behind embryo cryopreservation due to the poor survival, fertilization and success rates [73]. Currently, the literature reports oocyte vitrification yielding comparable outcomes to IVF with fresh oocytes in some cases [74].

The practice of oocyte cryopreservation became popular in Italy between 2004 and 2009 when legal restrictions permitted no more than three oocytes to be inseminated [75]. This clinical setting encouraged significant number of experiments on oocyte cryopreservation to be carried out in order to optimally preserve the supernumerary embryos retrieved in ovarian stimulation cycles. Nowadays, the primary indications for oocyte cryopreservation are for single women at risk of losing ovarian function due to oncology treatments, systemic illnesses and genetic syndromes. It is also widely used in centers with embryo donation programs to eliminate the need to synchronize the cycles of egg donors and recipients. Many women are aware of their fertility windows; requests for social egg cryopreservation to prolong the window of fertility are becoming more common. Oocyte cryopreservation is also applicable when there is an unexpected failure to obtain sperms on the day of oocyte retrieval in a fresh ART treatment cycle [73].

The pregnancy outcomes and health of the children following oocyte cryopreservation have been reported in literatures based on over a 1000 cases. Levi Setti et al. reported a higher rate of first trimester miscarriage in pregnancies following oocye cryopreservation in comparison to fresh cycles; [75] Oktay et al. looked at pregnancies following slow oocyte freezing and reported similar finding of increasing miscarriage rate [76]. Either in comparison to spontaneous conception or fresh ART cycles, reassuringly no differences have been reported in the rates of ectopic pregnancy and congenital anomalies [75, 77–79]. Levi Setti et al. reported a higher mean birth weights in singleton and twins following oocyte cryopreservation in comparison to fresh treatment, this pattern is similar to pregnancies following embryo cryopreservation [75]. Other studies, in contrast, did not see any difference in the mean birth weight. [77, 79] The study of Levi Setti et al. also recorded 138 pregnancies from 63 patients who had pregnancies in both fresh and thawed oocyte cycles; in these pregnancies, the miscarriage rate and mean birth weight were the same [75].

The advances in in vitro maturation (IVM) have enabled the development of the technique of immature oocyte cryopreservation. Immature oocyte cryopreservation is beneficial when a high proportion of oocytes retrieved are immature following ovarian stimulation. It is also potentially useful for patients who are not suited to undergo ovarian stimulation; one example of this group of patients is girls who are prepubertal. In a recent paper, it is stated that so far only one live birth following immature oocyte cryopreservation has been recorded in the literature [80]. Another experimental technique, which is increasingly common, is ovarian tissue cryopreservation. Similar to immature oocyte cryopreservation, ovarian tissue cryopreservation can also remove the need to perform ovarian stimulation for retrieving mature oocytes. Currently over 30 cases of live births following ovarian tissue transplants have been reported [80]. Histological analyses of harvested ovarian tissues have not detected any metastatic cancers for oncology patients undergoing fertility preservation [81, 82].