Microphthalmia and Optic Nerve Hypoplasia Induced by Maternal Coronavirus Infection?


 Possible mother-to-child transmission of severe acute respiratory syndrome-coronavirus type 2 (SARS-CoV-2) during pregnancy is still a matter of debate. We studied the impact of SARS-CoV-2 infection on 56 complete households, including 27 newborns whose mothers were pregnant when exposed to the virus. Three perinatal SARS-CoV-2 transmissions with mild symptoms in affected neonates were recorded (two cases confirmed by PCR, the third one based on clinical findings). In addition, we observed a severe eye malformation (unilateral microphthalmia, optic nerve hypoplasia, and congenital retinopathy) associated with maternal SARS-CoV-2 infection in weeks 5 and 6 of embryonic development. This embryopathy could not be explained by other infectious agents, genetic factors, or drug use during pregnancy. Eight other women with a history of SARS-CoV-2 infection prior to gestational week 12, however, delivered healthy infants.Conclusion: The repeated occurrence of mother-to-child transmission in our cohort with risks that remain incompletely understood, such as long-term effects and the possibility of an embryopathy, should sensitize researchers and stimulate further studies as well as strongly support COVID19 vaccination recommendations for pregnant women.Trial registration number: NCT04741412Date of registration: November 18, 2020


Introduction
Maternal infections with severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2) during pregnancy usually do not have severe consequences for mother and child. As some pregnant women, however, become seriously ill with coronavirus disease of 2019 (COVID-19), the primary risk to the infant appears to arise from maternal illness [1]. Transmission of SARS-CoV-2 to the fetus is probably a very rare event [2,3]. Therefore, possible complications of such pathogen spread are scarce and hard to capture. Potential harm to the fetus depends on many factors, particularly on the developmental stage of the unborn child at the time of viral infection.
We studied the impact of SARS-CoV-2 infection on 56 complete households, including 27 newborns whose mothers were pregnant when exposed to the virus. Among these infants, we recorded two unquestionable mother-to-child transmissions of SARS-CoV-2 shortly before or during delivery. In addition, the son of a woman with early gestation COVID-19 has an eye malformation similar to those observed after maternal rubella in the rst trimester of pregnancy, an issue to be put up for discussion in the scienti c community.

Study population
In a longitudinal study of 56 complete households (231 individuals) with one to eight members who had COVID-19 and/or developed antibodies against SARS-CoV-2, we investigated the course of disease, immune responses, and long-term consequences of the infection (ClinicalTrials.gov NCT04741412). This study was approved by the ethics committee of the University Erlangen-Nürnberg and conducted in accordance with the principles of the declaration of Helsinki. All individuals or their legal guardians provided written informed consent to participate. The cohort included 27 newborns of women who had acquired a SARS-CoV-2 infection during pregnancy. Clinical and sonographic follow-up examinations were performed in the rst 5 weeks after birth.

Diagnosis of SARS-CoV-2 infection
SARS-CoV-2 infection was diagnosed based on exposure history, clinical manifestation, and a positive reverse transcription-polymerase chain reaction (RT-PCR) result. Material from respiratory tract swabs, placental tissue, and cord blood samples were investigated. The qualitative cobas SARS-CoV-2 dualtarget RT-PCR assay (SARS-CoV-2 speci c target 1: ORF1/a region; pan-Sarbecovirus-speci c target 2: envelope E region), running on an automated PCR system (cobas 6800; Roche Diagnostics, Mannheim, Germany) displays cycle threshold (Ct) values for both viral target sequences and the internal control. According to the manufacturer's evaluation, the detection limit for SARS-CoV-2 RNA in respiratory tract swabs is 0.009 tissue culture infectious dose 50% (TCID50)/mL for target 1 and 0.003 TCID50/mL for target 2.
Flow cytometry-based assessments of antibodies against the SARS-CoV-2 spike protein were performed as described previously [4,5]. In brief: A plasmid encoding SARS-CoV-2 spike was co-transfected together with a green uorescent protein (GFP)-encoding plasmid into HEK293T cells. Two days later the cells were incubated with serum samples from patients (1:100 dilution), followed by staining with a secondary antibody mixture of PE-conjugated anti-human IgA (Southern Biotech, Birmingham, U.S.), AF647conjugated anti-human IgG (Southern Biotech) and DyLight405-conjugated anti-human IgM (Jackson ImmunoResearch, Ely, U.K.) antibodies. Stained cell populations were analyzed using a Gallios ow cytometer (Beckman-Coulter). The antibody TRES 224 [5,6] recognizing the SARS-CoV-2 spike protein served as positive control.
An electrochemiluminescence immunoassay to detect IgG speci c for the SARS-CoV-2 nucleocapsid was performed by SYNLAB International GmbH (Weiden, Germany).

DNA sequence analysis
After obtaining informed consent from the parents, genomic DNA of a newborn patient with an eye malformation was investigated; rst by Sanger sequencing of the gene PAX6, variants of which are most frequently associated with optic nerve hypoplasia. Speci c primers covering the exons and intron-exon boundaries were designed using the online design and analysis tool ExonPrimer (https://ihg.gsf.de/ihg/ExonPrimer.html) and the In-silico PCR tool from the University of California, Santa Cruz (https://genome.ucsc.edu/). Primer sequences and thermal cycling conditions are available upon request. DNA extraction, PCR, and sequencing were performed as described previously [7]. Secondly, a blood sample from this patient was sent to a provider of genetic diagnostics and sequencing services (CeGaT GmbH, Tübingen, Germany) for whole exome sequencing using the Illumina NovaSeq6000 Sequencing Systems. The bioinformatic data obtained were analyzed with the Golden Helix GenomeBrowse tool (Golden Helix, Bozeman, USA). Our selection of potentially relevant genes was based on the gene panels for optic nerve atrophy and ocular malformations from CeGaT (54 genes in total). Criteria for evaluating detected variants were for example population allele frequencies, genomic positions, predicted effects on biological function and data published in the available scienti c literature.
Each potentially pathogenic variant identi ed in this study was assessed with the mutation prediction tools Mutation Taster (Charite, Berlin, DE; Cardiff University, Cardiff, UK) and the Ensembl Variant Effect Predictor also containing the SIFT and PolyPhen-2 scores for protein changes (European Molecular Biology Laboratory's European Bioinformatics Institute, Hinxton, UK).

Results And Discussion
Clinical ndings in our study population indicate a very low risk of severe COVID-19 in children and adolescents. More than half of the individuals below 18 years of age remained asymptomatic (Table 1).
In 80% of the families investigated, an adult was the rst household member with COVID-19 symptoms. Speci c antibodies were detected in 69% of children and adolescents and in 88% of adults exposed to SARS-CoV-2 (Table 1). Apart from a 15-year-old girl who was hospitalized because of a coincidence of COVID-19 with manifestation of type 1 diabetes mellitus, no subject below 18 years had to be hospitalized.
In our cohort of pregnant woman, nine maternal SARS-CoV-2 infections occurred in the rst trimester of pregnancy. One of these mothers who had shown typical COVID-19 symptoms from day 54 after the last menstrual period (day 40 after conception) for one week, followed by an otherwise uneventful pregnancy, delivered a boy with unilateral microphthalmia, microcornea, and hypoplasia of both the optic nerve and the neurosensory retina (Figs. 1 A-C). Similar eye abnormalities, uni-or bilaterally, are seen in neonates with rubella embryopathy. The patient's second eye and visually evoked potentials under ash stimulation (Fig. 1 D) were found to be normal. Any relevant family history as well as rubella, toxoplasmosis, herpes or cytomegalovirus infection have been excluded; vaccination of the mother against rubella before becoming pregnant and protective levels of rubella-speci c IgG were evident. As the father had contracted COVID-19 ve days before his wife, SARS-CoV-2 replication in the mother most likely happened in weeks 5 and 6 of embryonic development, during oculogenesis [8]. Maternal PCR testing for SARS-CoV-2 (nasopharyngeal swab) three days after symptom onset was positive. Both parents developed speci c antibodies against the SARS-CoV-2 spike protein. At birth in gestational week 41 + 1, the infant weighed 4,000 grams; SARS-CoV-2 PCR tests were negative, while IgG against the nucleocapsid of SARS-CoV-2 (most likely of maternal origin) were still detectable in the cord blood.
Ultrasonography of brain and abdomen did not reveal additional malformations.
We attempted to rule out genetic causes of optic nerve hypoplasia in this patient by whole exome sequencing and careful analysis of 54 genes possibly involved in developmental anomalies of the eye (complete CeGaT panels for microphthalmia/coloboma and optic nerve atrophy, includingABCB6, BMP4, MCAT, OPA1, OPA3, PAX2, PAX6, RAX, THEM126A, andWFS1). We detected a total of three gene variants with an allele frequency below 1%: The rst of them, heterozygousABCB6variant c.649C > A (p.Q217K; transcript CCDS2436.1; ENST00000265316; NM_005689.4), was classi ed as benign by each of the mutation prediction tools used. It affects an exon in one of the two transcripts produced but an intron in the other transcript. The second possible mutation, heterozygousRAXvariant c.311G > A (p.G104D; transcript ENST00000256852.7) which is identical with c.565G > A (p.A189T) in anotherRAXtranscript (CCDS11972.1; ENST00000334889; NM_013435.3), does not, in our opinion, explain the phenotype, becauseRAXvariants have been considered as pathogenic in the literature only when present homozygously or compound-heterozygously. The third nding of potential signi cance, heterozygousMCATvariant c.235C > G (p.R79G; transcript CCDS14045.1; ENST00000327555; NM_014507.3), was classi ed as benign based on the SIFT and PolyPhen-2 scores and has been detected more frequently in a Swedish subpopulation.
Known teratogenic factors that may in uence optic nerve development, such as maternal alcohol consumption during early gestation, smoking, drug abuse, and maternal diabetes mellitus, were also excluded. As the patient's mother did not take any potentially teratogenic drugs, our ndings suggest the eye abnormalities of the infant to be due either to materno-fetal transmission of SARS-CoV-2 or to an indirect effect of the maternal SARS-CoV-2 infection at the time of optic tract development. This conclusion must be taken with caution, because the only link that connects the association is the fact that the patient's mother contracted COVID-19 during the relevant stage of embryogenesis. Given the high incidence of SARS-CoV-2 infection, the link may be entirely coincidental. On the other hand, this nding has to be reported to the scienti c community to allow further investigation and to raise awareness of rare complications associated with COVID-19.
In order to identify the pathogenetic mechanisms underlying a link between SARS-CoV-2 infection and ocular developmental anomalies, further studies involving animal experiments would be required. Even if a direct impact of the virus on neurodevelopment was considered unlikely, cytokine storm and hyperin ammatory processes in a pregnant woman with SARS-CoV-2 infection could still increase the risk of neurodevelopmental disorders in early stages of pregnancy [9]. Based on the biological properties of SARS-CoV-2, Leyser and colleagues speculate that the virus might neutralize the maternal immune response by affecting interferon type 1 expression, similar to Zika virus, thereby facilitating virus spread and increasing the risk of damage to progenitor cells of the fetal brain [9].
The other 8 newborns of mothers with SARS-CoV-2 infection during early gestation did not show any clinical or ultrasonographic abnormality.
In 23 of the 27 neonates investigated in this study, no morbidity related to preceding SARS-CoV-2 infection of their mothers was observed, whereas maternal infection at the very end of pregnancy (n = 5), before su cient passive immunity could be conferred to the baby, led to positive SARS-CoV-2 PCR results and mild COVID-19 symptoms in two newborns (40%; one male, one female), including a generalized postinfectious exanthema ( Supplementary Fig. 1 A) that lasted for up to 14 days. An almost identical skin rash was observed with three days delay also in the otherwise healthy, postnatally SARS-CoV-2 PCRnegative twin of the affected girl. Knowledge about such cutaneous manifestations of COVID-19 is still limited and will improve differential diagnosis. In one of the two symptomatic newborns, clearly elevated but continuously decreasing D-dimers were recorded ( Supplementary Fig. 1 B), explainable by an obvious cephalohematoma rather than by COVID-19-related thrombophilia.
Sonography of the brain and adrenal glands of the 5 perinatally exposed infants did not reveal any additional abnormalities.
The comparatively mild course of disease in our pediatric subpopulation con rms what has been observed by others [10], although current studies have not yet taken issues such as long COVID or myalgic encephalomyelitis/chronic fatigue syndrome [11] su ciently into account. In addition, the data were obtained prior to the occurrence of SARS-CoV-2 variants of concern which appear to be transmitted more easily [12]. Perinatal SARS-CoV-2 infections of the infant leading to a mild course of disease have been reported elsewhere, too [1,10], although their frequency in our study was higher than expected. An embryopathy associated with maternal COVID-19 is described here for the rst time. Both should sensitize researchers and stimulate more systematic studies. If further cases of potentially SARS-CoV-2induced malformations become known, COVID-19 vaccination recommendations for pregnant women may need to be reconsidered. This study was funded in part by the German-Swiss-Austrian ectodermal dysplasia patient organization.

Competing interests:
The authors declare that they have no con ict of interest.
Availability of data and material: The datasets used and analyzed during the study are available from the corresponding author upon reasonable request. Ethics approval: The study was approved by the ethics committee of the University Erlangen-Nürnberg (project code 212_20 B).

Consent to participate:
All adult participants provided written informed consent; in the case of minors, parental consent was obtained.

Consent to publication:
The parents of the boy shown in Figure 1A consented to publication of their son's photo.