First Trimester Pregnancy Outcomes in a Large ART Center From the Lombardy County (Italy) During the Peak COVID-19 Pandemic

Paolo Emanuele Levi-Setti (  paolo.levi_setti@humanitas.it ) Humanitas Research Hospital Federico Cirillo Humanitas Research Hospital Valentina Immediata Humanitas Research Hospital Emanuela Morenghi Humanitas Research Hospital Valentina Canevisio Humanitas Research Hospital Camilla Ronchetti Humanitas Research Hospital Annamaria Baggiani Humanitas Research Hospital Elena Albani Humanitas Research Hospital Pasquale Patrizio Yale University


Introduction
At the beginning of 2020 the Italian Lombardy region was hit by an "epidemic tsunami" with almost no similar incidences in other parts of the World before 1 . Based on the rst investigation on SARS-CoV-2 seroprevalence in one of the two initial lockdown areas in Lombardy, Lodi, at the beginning of the outbreak, the virus exposure was detected in 28% of asymptomatic blood donors 2 . The Lombardy region is home to 10 million inhabitants and accounts for 37% of cases and 53% of deaths of the country, as of April 15, 2020 2 , but the presented prevalence is likely to be underestimated by up to seven to ten times due to the lack of supplying of nasopharyngeal swabs during the pick period.Although o cially the rst coronavirus disease (COVID- 19) case was reported on February 21st at the Codogno (Lodi) Hospital located 57 km (about 35 miles) from our Fertility Center, a recent epidemiological analysis demonstrated that the new severe acute respiratory syndrome virus 2 (SARS-CoV-2) was present in the Northern Italy at least 3 months (end of December 2019) before the o cial recognition 3 . Therefore, it is highly plausible that during the months of January-February 2020, before the interruption of clinical services and the lockdown, several infertile patients, unrecognized because asymptomatic for COVID-19 signs, were treated with ART cycles and achieved pregnancies. Available data, albeit very limited, on the impact of the SARS-CoV-2 on second and third trimester pregnancies suggest an increased risk for preterm deliveries 4,5 and the risk for vertical transmission is minimal, although a possible signi cant higher stillbirth rate has been recently suggested 6 .
Case series have reported the detection of SARS-CoV-2 in different biological material, from semen to human breastmilk 7-10 , and the possible vertical transmission of the virus from an infected mother to her new born 11 raised serious concerns in the embryologist community worldwide [12][13][14] . However, there is very limited information about the possible consequences of SARS-CoV-2 infection on ART performance and early pregnancy outcome.
The aim of this work is to ll the gap by evaluating early pregnancy outcomes after assisted reproduction during the early peak time of COVID-19 in Lombardy before the region o cially ordered stopping all IVF treatments (with the exclusion of fertility preservation for oncologic patients and emergency or urgent procedures) and other treatments considered non-essential. Although the patients reported in this analysis were not positively identi ed as affected by COVID 19, the fact that they were treated during the unrecognized early peak of the pandemic make them a plausible exposed group and an interesting speci c patient population to assess the effect, if any, of the pandemic on early stages of pregnancy, and on the biochemical and rst trimester abortion rates 15 .

Material And Methods
This is a retrospective observational analysis from a single, tertiary care, university-a liated fertility centre, located in Lombardy, Rozzano, Milan, Italy, including all couples who underwent Assisted Reproduction (ART) procedures (fresh and frozen transfer cycles) during two time periods used for comparison: November 1 st , 2018 to February 28 th , 2019 (non COVID-19 risk) and November 1 st , 2019 to February 29 th , 2020 (COVID-19 risk.
No exclusion criteria were considered.
Patients were from all over Italy with a prevalence (70%) from the Lombardy region. The great majority of patients spent the entire period of the ovarian stimulation/monitoring in preparation for their treatments at 1-2 hours car distance from the hospital. Information collected included: female age, BMI and smoking habits, duration of infertility, basal Follicle Stimulating Hormone (FSH), Anti Mullerian Hormone (AMH), Antral Follicular Count (AFC), indication for ART treatment, primary or secondary infertility and previous abortions. In addition, follow up data about pregnancy, rst trimester abortion, and ectopic pregnancy outcome were collected until June 15, 2020.
No data about serological or oropharyngeal swab was considered. Of note, one embryologist with very mild symptoms had a con rmed positive oropharyngeal swab for SARS-CoV-2 during the study period, but no one else among medical, embryological, nurse and staff (about 60 people) had symptoms relatable to the viral infection.
Patient follow up included rst Beta-hCG's performed 12 days after blastocyst transfer or 14 days after cleavage stage transfer and for pregnant patients repeated every 48 hours, until reaching at least 1,500 UI/ml. Transvaginal ultrasound was scheduled 4 weeks after transfer or earlier in case of abdominal pain and vaginal bleeding or abnormal rising Beta hCG levels. All non-essential ART activities were interrupted on April 16, 2020. Only emergency services, including fertility preservation for oncological patients and pregnancy follow-ups were allowed during the lockdown period.
Patients were required to email the results of ultrasounds if performed in other facilities. Every day, medical staff and assistants on duty called patients and updated the pregnancy outcome in their medical charts. Less than 5% of patients were lost at the follow-up that was considered concluded only when all known pregnant patients completed at least the 12 th gestational week (as of June 15, 2020).

Statistical Analysis and variable description
Clinical pregnancy was de ned as a pregnancy diagnosed by ultrasonographic visualization of one or more gestational sacs or or de nitive clinical signs of pregnancy 16 . It therefore included ectopic pregnancy. Pregnancies with bHCG levels reaching less than 1,000 mIU/ml, and after exclusion of ectopic localization, were considered biochemical.
Data were described as number and percentage, or mean and standard deviation, as appropriated.
Associations with period (pre and during COVID) were explored with χ2 test for categorical variables, or t student test for Gaussian continuous variables, or Mann Whitney for asymmetrical continuous variables. Association with abortion rate was explored with logistic regression analysis. Independent variable with a p value under 0.2 were then submitted to a multivariable logistic regression analysis. Pre and during COVID period variables were included in the multivariable analysis being the primary endpoint of the research. A p value under 0.05 was considered as signi cant. All analyses were conducted using Stata Statistical Software: Release15 (College Station, TX: StataCorp LLC).

Ethical approval
Generally, all patients undergoing ART procedures consent in writing that their medical records can be used for research purposes if anonymity and con dentiality is protected. Since both conditions were met, this study had expedited review and approval by the center's Institutional Review Board (IRB), Humanitas Clinical Institute Ethic Committee.
All methods were performed in accordance with the relevant guidelines and regulations and the present research has been performed in accordance with the Declaration of Helsinki.
An informed speci c consent was obtained from all participants and/or their legal guardians.
The study was also approved by our Independent Ethical Committee on May 14 th , 2020 (protocol n. 37/20).
Female age was 37.11 ± 4.17 in non COVID and 36.56 ± 4.19 in the COVID-19 period (p=0.026). No differences were found in other patient's demographic, ovarian response, and laboratory parameters (see  Table III).
A multivariate analysis investigating differences in the abortion rate (Table IV)

Discussion
The World Health Organization on 11 th March 2020 declared the pandemic status for COVID-19, but Italy, mostly in the northern region of Lombardy, had already begun experiencing the severity of the COVID-19 since February 21, 2020, or even before. To prevent the diffusion of contagion, to avoid overwhelming the healthcare system, as well as reduce the worries of establishing a pregnancy during uncertain times, on There is still a paucity of information about the impact, if any, of COVID 19 infection on early pregnancy ( rst trimester) outcome. Since it is very plausible that the novel SARS-COV2 virus could have spread, unrecognized, in Lombardy area already few months before ART treatments were suspended, this study investigated whether the several infertile asymptomatic patients who were treated and achieved pregnancies during that at risk time-period had a different pregnancy outcome when compared to a similar infertile population dataset taken from the same season but one year before (non-COVID-19 risk).
The results of our analysis showed no differences in early pregnancy outcomes between the controls and the COVID-19 risk exposure in terms of implantation, pregnancy, biochemical and abortion rate both in fresh and frozen ART cycles. The lack of an increased risk for miscarriage or other adverse outcome in early pregnancies obtained during high risk COVID-19 exposure is reassuring.
The strength of the present study relays on the large sample analyzed in a geographical area considered the epicenter of the pandemic in Europe and where the incidence of SARS-COV-2 positive cases was very high in general asymptomatic population. SARS and MERS epidemics showed no correlation with fetal malformations. However, the clinical course of COVID-19 disease and the response to treatments seem to differ from other previous types of coronaviruses 19 . In order to fully understand pathogenesis and epidemiology of SARS-CoV-2 during pregnancy, further research is needed focusing on the time of maternal infection, gestational age, role of comorbidity factors, and adverse outcomes. Luckily, preliminary studies on pregnant women infected with SARS-CoV-2 give an optimistic outlook regarding the clinical course 20 .
During the rst months of the pandemic, the American Society for Reproductive Medicine Coronavirus/COVID-19 Task Force reviewed ninety-seven articles about pregnancy and coronavirus in order to clarify the effect of the novel virus on human reproduction and pregnancy 8 . At the time, no reported studies did examine pregnant COVID-19 patients at earlier stages of pregnancy. Few data regarding the effect of SARS-CoV-2 on human reproduction are available because the virus is novel and has only recently infected humans. The SARS-CoV-2 virus enter into human cells using ACE2 receptors. The reproductive system in men expresses ACE2 in Leydig cells in the testis and it may play a role in spermatogenesis. Gonadotropin-dependent expression of ACE2 has been reported also in human female gonads 21,22 , but it is still unknown if the SARS-CoV-2 virus uses ACE2 receptors in the human reproductive system and what, if any, impact this might have on oocyte quality, embryo development, or the consequent pregnancy. Our data is an indirect con rmation (since we did not show presence or absence of the virus in our asymptomatic pregnant ART patient population) of existing evidence from two case-control studies (involving 46 patients and 287 controls) showing that COVID-19 during early pregnancy is not more severe than among non-pregnant women 23,24 .
The real impact that COVID-19 infection can cause on fertility and human reproduction remain obscure. Despite the overwhelming magnitude of the disease and its worldwide prevalence, information regarding the effects of the novel coronavirus on human reproduction are currently limited. This lack of evidence should not be considered reassuring because less than 1 year have elapsed since the novel coronavirus jumped species and infected humans.
It is important to study other possible long-term effects on male and female gametes , speci cally whether there might be shedding of virus in some individuals that might even affect the safety and storage of gametes 18 . Evidence continues to emerge regarding effects of the novel coronavirus in pregnancy and some initial reports suggest that complications, particularly after delivery, may be increased 9 , even if outcomes for infants are largely reassuring when considering potential effects of SARS-CoV-2 infection acquired before or during birth 5,25 . Further studies are necessary, and additional data regarding outcomes of early pregnancies in demonstrated infected pregnant women should be collected 26 .
The main limitation of this study is the retrospective nature of the analysis and the high plausibility (but not corroborated) of the COVID-19 exposure since the ART treatments occurred during the early peak of the Italian pandemic.

Conclusion
Our preliminary dataset mitigated concerns for negative reproductive consequences from COVID-19 pandemic and demonstrated no increased risk for miscarriage during the rst trimester of ART pregnancies achieved during the emergence of the COVID 19 health crisis.

Declarations
Authorship contribution PELS, FC, VI and CR, designed the study, collected data, and drafted the manuscript. VC, AB and EA collected clinical data. EM analyzed clinical data and performed the statistical analysis. PP analyzed data and revised the manuscript. All authors participated to critical discussion and nal manuscript. Tables   Table I. Patients, demographics, indications for ART and cycle characteristics for the total fresh cycles analyzed in the non COVID and COVID-19 times frame.