COL4A1 gene mutations and perinatal intracranial hemorrhage in neonates: case reports and literature review

Intracranial hemorrhage may represent a complication of the perinatal period that affects neonatal morbidity and mortality. Very poor data exist about a possible association between mutations of the type IV collagen a1 chain (COL4A1) gene and the development of intracranial hemorrhage, and only sporadic reports focus on intracerebral bleedings already developing in utero or in the neonatal period in infants with such a mutation. This study presents a case series of term neonates affected by intracranial hemorrhage, with no apparent risk factors for the development of this condition, who were carriers of COL4A1 gene variants. This study also provides a review of the most recent scientific literature on this topic, specifically focusing on the available scientific data dealing with the perinatal period.


Introduction
Intracranial hemorrhage (ICH) develops most often in premature newborns, while its occurrence is relatively uncommon and may represent an incidental finding in term neonates (1,2).Thanks to increased clinical awareness and improved neuroimaging techniques, however, ICH at term has been diagnosed more often in recent years (1,3).In some cases, the clinical impact of such a condition may deeply influence the perinatal course and affect neonatal morbidity and mortality (4)(5)(6)(7).ICH may develop in utero or after birth.The rates of fetal ICH are 0.46 per 1,000 deliveries and 0.9 per 1,000 pregnancies at referral centers (8).In full-term infants, possible risk factors for the development of perinatal cerebral bleeding include infections, malformations, asphyxia, and thrombophilia (4,5,7,9).Besides these conditions, familial susceptibility with a genetic predisposition to a hereditary form of porencephaly may play a role, but available data are still inconsistent.
Type IV collagen a1 chain (COL4A1) gene mutations are responsible for a hereditary autosomal dominant cerebrovascular disease, characterized by a wide phenotypical spectrum, which may include the development of ICH.To date, only a relatively small number of neonates carrying COL4A1 gene mutations and developing ICHs has been reported (10)(11)(12)(13)(14)(15).However, the diagnosis of COL4A1 gene mutations is probably underestimated because of the failure to search for the mutation because of its heterogeneous clinical presentation.When focusing on the perinatal period, the data available are even more infrequent, and the exact clinical features of such disorders are still far from clear.
The aim of this study was to describe the case of two term neonates affected by ICH of apparently unclear origin who showed mutations of the COL4A1 gene, expanding the spectrum of perinatal disease attributable to COL4A1 mutations and adding information in a field where knowledge is still far from clear.This study also provides a review of the currently available scientific literature on this topic, specifically focusing on the perinatal period.

Case reports
We describe the clinical history of two neonates with apparently unexplainable ICH.A COL4A1 gene mutation was detected in both patients.

Patient 1
Patient 1 (Table 1) was a male neonate born at a gestational age (GA) of 40 weeks by spontaneous delivery, with a birth weight (BW) of 2,900 g.One month before delivery, microcephaly and polihydramnios were detected.Amniocentesis was normal and maternal TORCH assessment was negative for recent infections.
At birth, the neonate showed normal cardiorespiratory adaptation (Apgar score at 1 min: 9; 5 min: 10).Cerebral ultrasound (CUS) could not be performed because of the reduced diameter of the front cranial fontanel.Electroencephalography (EEG), performed on day 6 of life, was normal.Magnetic resonance imaging (MRI) of the brain, performed on day 9 of life, showed a volumetric reduction of the brain with a simplification of cortical convolutions, polycystic glio-malacic lesions located in the left parietal and occipital corticalsubcortical region and in the right parietal-mesial region, hemosiderin deposits suggestive of previous hemorrhage in the right posterior paraventricular region, asymmetry of the posterior horns of the lateral ventricles (right > left), and reduced thickness of the intermediate segment and of the splenium of the corpus callosum.Diffusion-weighted imaging (DWI) did not highlight any lesion suggestive of recent ischemic damage (Figure 1).On the day 17 of life, the neonate developed an episode of diffuse muscular hypertonus and gaze deviation.The EEG performed on this occasion showed right focal specific lesions and therapy with barbiturate was therefore begun.The subsequent EEG was normal.
After the first month of life, the infant developed muscular hypertonus of the four limbs, axial hypotonia, reduced reactivity, incomplete Moro reflex, right head deviation, and dysphagia.
The infant's audiological and ocular assessments were normal.Renal function was normal.Thrombophilia testing was performed; prothrombin time, partial thromboplastin time, bleeding time, fibrinogen, antithrombin III protein C, protein S, activated protein C (APC) resistance, homocysteinemia, and testing for mutations in the prothrombin and factor V were all normal.A heterozygosity of the methyltetrahydrofolate reductase (MTHFR) gene (C677T) was detected.Immunoglobulins for rubeola, herpes viruses 1/2, Toxoplasma gondii, cytomegalovirus, and urinary polymerase chain reaction for cytomegalovirus excluded viral infections.
After the proband's parents provided written informed consent, molecular analyses of COL4A1 were performed using a custom panel [NimbleGen SeqCap Exome Enrichment Kit (Roche Sequencing Solution, CA, USA)] according to the manufacturer's protocol and sequenced on the Illumina platform.Annotated data were filtered out to exclude intronic variants, synonymous variants not predicted to affect splice sites, and variants with minor allele frequency (MAF) of ≥0.1% in the Genome Aggregation Database (gnomAD).Trio analysis identified the heterozygous missense variant c.2705C>G (p.Pro902Arg) of the COL4A1 gene (NM_001845) in maternal segregation in our patient.
The COL4A1 gene is an autosomal dominant gene of susceptibility to cerebral hemorrhages.The variant was not known in the scientific literature and was reported in the reference population database (gnomAD AF: 0.0004351; dbSNP: rs146134172; ClinVar ID: 80023).At the comparative genomic hybridization (CGH) array, the infant was also positive for the microduplication 16q24.3 with paternal segregation.
The copy number variants (CNVs) of region 16q24.3are frequently associated with KBG syndrome.KBG syndrome is often caused by mutations in the AKNRD11 gene, microdeletions of 16q24.3,and, less frequently, by microduplications of this same region (16).Furthermore, the patient did not have phenotypic features peculiar to KBG syndrome, such as triangular face,  hypertelorism, brachycephaly, synophyrys, prominent nasal bridge, and elongated philtrum.Furthermore, he had never experienced a single seizure episode that then resolved.Therefore, this diagnosis was excluded.As integration, the analysis of genes associated with microcephaly was performed and was negative.
The patient received rehabilitative physiotherapy during hospitalization and after discharge as well as speech therapy support.Percutaneous endoscopic gastrostomy and jejunostomy were performed because of persistent dysphagia, feeding difficulties, and gastroesophageal reflux.A follow-up at 7 years of age showed that the infant developed severe neurodevelopmental delay, spastic tetraparesis, and relational and language disorders.

Patient 2
Patient 2 (Table 1) was a male neonate born at a GA of 38 weeks by spontaneous delivery, with a BW of 3,870 g and Apgar scores of 9 and 10 at 1 and 5 min, respectively.Pregnancy was uneventful except for placental dysfunction during the first trimester, which was treated with aspirin.The early postnatal period was normal.On day 7 of life, the neonate developed jaundice, multiple vomiting episodes, and irritability.CUS showed asymmetric dilatation of the lateral ventricles with moderate (grade II-III) bilateral intraventricular hemorrhage (IVH) and dilatation of the third ventricle.Brain MRI confirmed these findings and highlighted further hemorrhage in the context of the posterior cranial fossa and small parenchymal hemorrhage in the right cerebellar hemisphere, in the cisterns of the skull base, and in the left hemispheric extra-axial site (Figure 2).EEG showed electric seizures on the left occipital hemisphere, which were treated with barbiturate.
As a result of worsening dilatation of the cerebral ventricles, a ventricular drain connected to a subcutaneous reservoir was inserted 1 week later, which was then replaced with a ventriculo-peritoneal shunt.The patient's ocular and audiological assessments were After the proband's parents provided written informed consent, a molecular analysis of COL4A1 was performed using the custom panel with the same technique as described in the first case.Trio analysis identified heterozygous missense variant c.413C>G (p.Pro138Arg) of the COL4A1 gene (NM_001845) present in the newborn and in his mother.The variant was neither known in the scientific literature nor reported in the reference population database (gnomAD, HGMD).In silico bioinformatic prediction tools indicated that the variant was pathogenic.

Literature review
To review the literature about COL4A1 and perinatal intracranial hemorrhage in neonates, an extensive literature search in the MEDLINE database (via PubMed) was performed up to the year 2000.The keywords "COL4A1" AND "perinatal intracranial hemorrhage" OR "neonatal intracranial hemorrhage" were also searched as entry terms.All retrieved articles were screened, and then full texts of records deemed eligible for inclusion were assessed.References in the relevant papers were also reviewed and further articles were added if necessary.Papers written in languages other than English were excluded.
In Table 2, we systemically collected and summarized information on patients' characteristics, age at diagnosis, and genetic diagnosis, and we compared them to our case.
A total of 33 cases were reported in the literature before our case.In 25 of the 33 cases, the onset of ICH was prenatal, and pregnancy was terminated in 8 cases.

Discussion
COL4A1 and type IV collagen a2 chain (COL4A2) form the (a1)2(a2) IV heterotrimers and represent key components of basement membranes (28-31).COL4A1 is detectable ubiquitously in all tissues, including the vasculature, renal glomeruli, ocular structures, and muscles, and plays a role in the development of basement membrane during embryogenesis, in the cohesiveness of these membranes, and in the maintenance of vascular tone (32)(33)(34).The COL4A1 gene is located on the telomeric region of 13q (13q34) and is made up of 52 exons (35).Mutations of the COL4A1 gene may lead to a wide range of abnormalities variably affecting the brain (including the development of ICHs), retinal vasculature, ocular structures, renal glomeruli, and muscles, but no clear genotypephenotype association has been identified yet, even within the same family (15).COL4A1 mutations are inherited as an autosomal dominant trait and have near 100% penetrance with variable expression; de novo mutations or low-level parental mosaicism have been described (25).
Neonates described in the present case series were both born at term after pregnancies with no history of infections, traumatic injuries, or thrombotic events and had a normal cardiorespiratory adaptation at birth.The heterozygosity of the MTHFR gene was highlighted in both cases, but no other anomalies were detected concerning the coagulative tests, platelet counts, and thrombophilia screening.ICH was diagnosed postnatally in both cases, although the radiologic findings of patient 1 showed the presence of glio-malacic lesions that were presumably of antenatal origin; even if it was not possible to assess the exact onset of the ventricular dilation detected in patient 2, the entity of the lesion was probably consistent with a prenatal origin.In our cases, prenatal MRI was not performed because of the late detection of anomalies during pregnancy in the first case and the low risk of pregnancy in the second case.However, as recently described by George et al. (36), prenatal COL4A1/A2 variations can show a range of ICH affecting the frontal lobes and basal ganglia, such as minor and/or unifocal ICH, as well as multifocal and bilateral lesions.In all cases of prenatal ICH, genetic testing for COL4A1/A2 variations should be taken into consideration because of the wide range of severity observed on fetal brain MRI.However, both patients came to our observation in the postnatal period; therefore, we had no other information regarding the decision-making process that took place during pregnancy.
The neurologic outcome was heterogeneous since it ranged from normal neurodevelopment (patient 2) to severely impaired neurologic development (patient 1).Small parenchymal hemorrhage in the right cerebellar hemisphere (black arrows).To date, only a few authors have investigated the role of COL4A1 mutations in the pathophysiologic mechanisms leading to intracerebral bleeding, and most of them focused on pediatric/adult patients rather than on neonates (37-39).Vahedi et al. found that some COL4A1 gene mutations may remain completely asymptomatic for years with no vascular changes on brain MRI, suggesting that the clinical expression may be heterogeneous even within the same family (37).Such findings are in agreement with subsequent data showing that brain MRI is abnormal in the majority of COL4A1-mutated patients (even in asymptomatic ones), but normal brain MRI may also be detectable in some mutation carriers with no brain manifestations (39).
Although there has been increasing interest regarding this condition in recent years, clinical experience about COL4A1related perinatal ICHs only results from sporadic case reports (Table 1).Gould et al. were the first to demonstrate in both a mice model and in humans that mutations in the COL4A1 gene may compromise the vascular basement membrane and lead to perinatal ICH (10).Since then, other authors reported the association between COL4A1 gene mutation and perinatal ICHs, but no exact genotype-phenotype association has been identified yet.
The involvement of the central nervous system in the presence of COL4A1 gene mutations varies widely from case to case in terms of onset timing, injury entity, and location (17-27, 40, 41).Features suggestive of COL4A1 mutations include severe and/or multifocal hemorrhagic or ischemic-hemorrhagic cerebral lesions, especially when such injuries are associated with porencephaly (27).
A deeper understanding of the relationship between COL4A1 mutations and the development of perinatal porencephaly may have important implications for disease prevention and optimization of perinatal care.Some authors suggested a possible interaction between environmental triggers and the genome (10).Although some neonatal cases were diagnosed after birth (as in our cases), the postnatal neuroimaging features were mostly consistent with an antenatal onset of ICH (Table 1).
It is crucial to emphasize the importance of the motherplacental-fetal triad in the developing of illness processes that cause brain abnormalities in fetuses or newborns.Geneenvironment interactions affect the brain development of newborns, children, and the mother-placental-fetal triad in both the short and long term.These interactions start at conception (40).During the initial 1,000 days of brain maturation, critical/ sensitive periods are more likely to result in lifelong developmental neuroplasticity (40).The hazardous stressor interaction between internal and external sources modifies the neural exposome by means of maladaptive developmental neuroplasticity.The first 1,000 days are when epilepsy and developmental problems mostly manifest (41).
Prenatal environmental factors, such as premature uterine contractions or other traumatic injuries, may activate some triggering mechanisms leading to vascular damage in at-risk patients (17, 42).It was hypothesized that specific preventive strategies, such as cesarean delivery instead of spontaneous delivery in at-risk individuals, might decrease the stress on the abnormal brain vasculature and, therefore, the severity of ICH in high-risk populations (10); however, further studies are needed.Nevertheless, according to the literature, perinatal ICHs occurred in neonates both by spontaneous and cesarean delivery, and brain injury has sometimes already occurred during pregnancy (Table 2).Finally, the neonates described in the literature were born either at term or prematurely (Table 2), and it might be put forward that some COL4A1 mutations may carry a higher risk for preterm birth and/or higher susceptibility to environmental factors than others (17).
Regarding the importance of prevention, it is also important to highlight that many children with developmental disabilities live in low-resource countries low or middle income country (LMICs) or high-income desert countries (HICMDs).
In these places, it is even more important to implement strategies aimed at identifying at-risk pregnancies as early as possible to provide families with accurate diagnoses and subsequent effective interventions.To this end, a biosocial model was developed that emphasizes women's reproductive health with trimester-specific maternal and pediatric health interactions (43).
The ability to perform genetic, neurophysiological, and neuroimaging assessments improves clinical decision-making for more effective interventions before pathological expression.
Support for the family, especially in the case of neurological pathologies in which we know how fundamental the first 1,000 days of life are, continues in the postnatal period (40).
Synergy between obstetric and pediatric healthcare providers can reduce neurological morbidities (43).
In conclusion, mutations of the COL4A1 gene as a potential risk factor for perinatal ICH are probably underestimated because of low clinical awareness, poor available data, and wide phenotypical spectrum; COL4A1 mutations should be suspected when otherwise unexplainable ICHs occur prenatally or in the neonatal period; if COL4A1 carriers are identified, familiar genetic counseling is advisable (37, 44); and genetic testing can be performed both prenatally and postnatally (24).There is still a lack of long-term follow-up data about the risk of ICH recurrence as long-term neurologic outcomes are missing.The current gap in knowledge in this field represents a call to action for clinicians/researchers and deserves further investigations.

FIGURE 1 (
FIGURE 1 (A,B) Encephalomalacia secondary to parenchymal hemorrhage: axial T2-weighted image shows the limited area of parietal encephalomalacia (black arrow) with ex vacuo enlargement of the lateral ventricle.(C) Corpus callosum thickness reduction (black arrow).

TABLE 1
Neonatal clinical features.

TABLE 2
Perinatal intracerebral hemorrhage associated with COL4A1 gene mutations.