Maternal complications following open and fetoscopic fetal surgery: A systematic review and meta‐analysis

Abstract Objective To establish maternal complication rates for fetoscopic or open fetal surgery. Methods We conducted a systematic literature review for studies of fetoscopic or open fetal surgery performed since 1990, recording maternal complications during fetal surgery, the remainder of pregnancy, delivery, and after the index pregnancy. Results One hundred sixty‐six studies were included, reporting outcomes for open fetal (n = 1193 patients) and fetoscopic surgery (n = 9403 patients). No maternal deaths were reported. The risk of any maternal complication in the index pregnancy was 20.9% (95%CI, 15.22‐27.13) for open fetal and 6.2% (95%CI, 4.93‐7.49) for fetoscopic surgery. For severe maternal complications (grades III to V Clavien‐Dindo classification of surgical complications), the risk was 4.5% (95% CI 3.24‐5.98) for open fetal and 1.7% (95% CI, 1.19‐2.20) for fetoscopic surgery. In subsequent pregnancies, open fetal surgery increased the risk of preterm birth but not uterine dehiscence or rupture. Nearly one quarter of reviewed studies (n = 175, 23.3%) was excluded for failing to report the presence or absence of maternal complications. Conclusions Maternal complications occur in 6.2% fetoscopic and 20.9% open fetal surgeries, with serious maternal complications in 1.7% fetoscopic and 4.5% open procedures. Reporting of maternal complications is variable. To properly quantify maternal risks, outcomes should be reported consistently across all fetal surgery studies.

remainder of the index pregnancy, potentially during any future pregnancies and throughout the woman's entire life. Fetal surgery offers no direct medical benefit to the mother, and from an ethical perspective, maternal risks should be minor and acceptable to the mother and family. 5 Information regarding safety of surgery is important for counselling and informed decision making; however, robust data on maternal complications of fetal surgery are lacking. One single-centre study of maternal outcomes following both open fetal and fetoscopic surgery performed between 1989 and 2003 found a number of short-term morbidities. 6

| Protocol and registration
This systematic review was conducted in accordance with preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidance. 8 The protocol was registered with the international prospective register of systematic reviews (PROSPERO-CRD42017082411).

| Eligibility criteria
All randomised, cohort, and case-controlled studies and case series reporting the results of open fetal or fetoscopic fetal surgery in humans from January 1990 to October 2018 were considered eligible.
No language restrictions were applied. Systematic reviews, narrative review articles, and case reports were excluded. There is no accepted numerical definition of a case series. 9 We used an empirical cut-off of at least three cases because of the rarity of some procedures and conditions searched for.

| Study selection
Two authors (A.S. and L.VdV.) screened all titles and abstracts independently, excluded irrelevant studies and then independently assessed the remaining full-text articles for eligibility; disagreements were resolved by consensus. Studies were excluded if the full text was unavailable online and the abstract contained insufficient information. Studies with interventions which were not fully described or were performed on the neonate instead of the fetus were excluded.
Interventions involving access to the uterus using a device with a total outer diameter of less than 1.5 mm were excluded; this cut-off was chosen to avoid procedures performed with needles only (eg, amniocentesis, fetal blood transfusion, thoracocentesis or vesicocentesis).
Studies of shunting were only included if the outer shunt diameter was greater than or equal to 1.5 mm or the shunt was inserted fetoscopically. Studies which did not report maternal outcomes were excluded. For the purpose of this study, preterm rupture of membranes (PROM), chorionic membrane separation (CMS), preterm labour, preterm delivery, and gestational age at delivery, although relevant, were not considered to be maternal complications. Studies from which data could not be extracted (eg, composite or combined outcomes given) and studies containing patient cohorts which appeared to have been published previously by the same authors were excluded.
What's already known about this topic?
• Fetal surgery, both open and fetoscopic, is now widely performed.
• Fetoscopy is perceived as safe for the mother, although specific data on maternal complications is lacking.
• Open fetal surgery is known to cause maternal morbidity, but the exact nature and frequency of complications is not well established across different centres and types of surgery.
What does this study add? •

| Maternal complications in the index pregnancy -postoperative
One study on laser photocoagulation for TTTS (n = 132) 110 Table 5.  3.9 | Maternal outcomes following the index pregnancy (long-term) In almost all studies of fetal surgery reviewed, long-term maternal follow up was not described. The seven studies that did so had a wide variation in the parameters described. Fertility does not appear to be negatively affected by fetal surgery, with the rates of de novo difficulties for conceiving in this review (3.81% following open fetal surgery and none following fetoscopic surgery) being comparable, if not less, than published rates of secondary infertility in the general population. 180 Similarly, the rates of miscarriage described (19.85% following open fetal and 13.67% following fetoscopic surgery) are similar to rates of spontaneous miscarriage in women who have not undergone fetal surgery. [181][182][183] Epidemiological studies 184 have suggested a worldwide preterm birth rate of 11.1% with a rate of 8.6% in "developed regions." 184 In the United States and United Kingdom, it is estimated at 9.8% 185 and 7.3%, 186 respectively. The preterm birth rate in this review following open fetal surgery (20.49%) is higher than the usual prevalence, but not higher following fetoscopic surgery (2.12%). Open fetal surgery was followed by uterine rupture or dehiscence in 6.89% and 11.09% of subsequent pregnancies, respectively, which is in line with published rates of rupture (6.2%) and dehiscence (12.5%) following a classical caesarean section. 187 Conversely, no uterine ruptures were reported following fetoscopic surgery.

| Overall maternal complication rates
This study included the commonest fetal procedures and, from a maternal perspective, involved similar surgical manipulations yet variable operating times. We included studies from multiple centres worldwide and attempted to identify the non-English literature. It is therefore likely that these results are generalisable to fetal surgery performed outside the included studies. An obvious weakness of this systematic review is that most studies did not include a control group.
Furthermore, we decided to pool data for meta-analysis despite having high heterogeneity in some results. Another weakness is the extraction of patient data from papers, which is prone to error given the variable reporting; it is possible that some patients had more than one complication and this was not noted or cumulative rates were as a consequence miscalculated.
This systematic review has identified a significant rate of maternal complications, which should be discussed with patients before embarking on fetal surgery. Large studies allow an estimation of the likelihood of these events, insomuch as the cases in these series are unselected and consecutive. Our systematic review search strategy may have missed relevant yet rare complications. For example, a letter to a journal editor describing maternal convulsions during general anaesthesia 188 was excluded as a case report according to our criteria.
In this circumstance, it appears that the patient was also part of the cohort of a study that was included, 47 but it is possible that other rare events published as case reports have been missed. An international, prospective registry of fetal and fetoscopic surgery, such as the Eurofoetus 189 and NAFTNet 190 registries, would be the best way to accurately determine complication types and rates and avoid missing rare complications.

| CONCLUSION
The maternal risks of fetal surgery are accepted by many patients and health care professionals for the possible benefit to the fetus. 191,192 This systematic review finds that studies of fetal surgery focus on the fetal outcomes of the procedure, and many fail to describe maternal complications. Fetal surgery comes at a risk to the mother, which may be underestimated by fetal therapists because of under-reporting and variable reporting quality. In order to properly quantify maternal risks, outcomes should be reported consistently across all studies of fetal surgery, preferentially in prospective registries.

CONFLICT OF INTEREST
All authors report no conflict of interest.

FUNDING INFORMATION
This research is funded by the Wellcome Trust (WT101957) and Engineering