The Epidemiology of Moebius Syndrome

Background: The epidemiology of Moebius Syndrome (MBS) is actually dicult to assess. In the present study we investigated the epidemiology of MBS in a well-dened population over a precise geographical area. Materials and Methods: Our University Hospital is the only national referral Centre for the diagnosis and treatment of MBS. Participants to this cross sectional study are patients affected by MBS who have been periodically followed by our medical staff since 1998. Most of the patients were referred to our hospital by the Italian Association of Moebius Syndrome (AISMO). Demographic data necessary for our purposes were made available by AISMO database updated to April 2018. Subjects were assigned to the geographical macro- areas which are the ones conventionally used for surveys and epidemiological investigations by the Italian National Institute of Statistics. Rates and prevalence of the MBS cases were calculated referring to the last survey of the Italian population available. Every study parameter was then calculated by reference to the whole country and to macro-area partition. Gender rate and the corresponding prevalence were calculated with respect to the weighted whole population and to the respective gender population. Chi-square analysis was adopted to investigate possible differences among geographical regions and/or gender. A p value <0.05 was considered statistically signicant. Results: One-hundred and sixty-four out of 212 MBS patients fullled our inclusion criteria. All cases were Caucasian and sporadic. The median age at diagnosis was 3.6 years, ranging from 0-55; this range was signicantly reduced to 0-5 years (median age at diagnosis: 2.2 years) for patients included after 2007. The birth prevalence calculated was 0.06 cases per 10000 live births with an overall prevalence of 0.27/100000 without any gender or geographical predisposition. Conclusions: The rate of MBS prevalence herein observed, rounded for possible underestimation, is 0.3/100000 people without any regional difference in the distribution of the cases. Our data conrm on a national basis the rarity of the disease.


Introduction
Moebius syndrome (MBS) is one of the most common congenital cranial dysinnervation disorder (CCDD). Its clinical features are impaired ocular motility, lagophthalmos, and lack of facial expression; these features are related to congenital non-progressive 6 th and 7 th nerve palsies that typically affect newborns bilaterally. MBS is diagnosed according to the "Bethesda Diagnostic Criteria", which have been recently updated with genetic testing aimed to ascertain diagnosis [1][2][3][4] . Patients who do not meet these criteria must be labelled as "Moebius-like" and considered as affected by a separate congenital disorder. This is of particular importance since MBS overlap many clinical features of other CCDDs with a well described genetical basis, such as congenital brosis of the extraocular muscles (CFEOM), Duane's syndrome, horizontal gaze palsy with progressive scoliosis (HGPPS), Poland syndrome and primary myopathic disorders. 5 More than a century after the rst description of the disease, 6 the etiology of MBS is still unclear; recent studies have postulated a multifactorial pathogenesis in which a fetal toxic exposure acts on a genetic predisposition responsible for vascular terminal instability and focal microcirculatory failure at the level of the lower brainstem. [7][8][9] However, it is not clear what causes these changes and why they speci cally disrupt development of the 6 th and 7 th cranial nerve nuclei; even less is known about the causes of the extra-ophthalmological signs and symptoms associated with MBS (e.g., lingual and palate dysfunction, hypoplasia of the hand, clubfoot, and thoracic abnormalities). The exact incidence and prevalence of MBS are not clear; the syndrome is considered a "rare disease" affecting a very small number of people. Clinicians and researchers estimate that this condition affects 1 in 50,000 to 1 in 500,000 newborns, but this estimate is based only on their personal experience with MBS patients, with no epidemiological basis. 2,10-12 In a Dutch series, the estimated prevalence of MBS was 0.002% of births (4 per 189,000 newborns); this evidence was obtained in the 1996, namely without the present diagnostic criteria for MBS. 13 The Orphanet Report Series, Rare Disease Collection 2019 reports the estimated prevalence\incidence per 100,000 as "unknown" with only 300 cases described in the literature. 14 Other epidemiological estimates of series reported worldwide are anecdotal, with no statistical basis. It is di cult to plan an epidemiological study of MBS for many reasons: 1) despite the new diagnostic criteria, the disease is often over-and misdiagnosed in newborns; 2) like many other rare diseases, MBS does not have a regional register from which to derive data for epidemiological purposes; 3) there are no referral centres that provide multidisciplinary care with consequent dispersion of cases; 4) few physicians have expertise in MBS and they may be di cult to reach; and 5) MBS, like other genetic disorders may carry social stigma leading affected individuals to self-marginalize. This study reports the epidemiology of MBS in a well-de ned population over a very long period; furthermore, we investigated whether there are geographical differences in MBS incidence\prevalence to identify factors that may cause or contribute to its appearance.

Methods
Since 1998, the University Hospital of Parma has been identi ed by the Italian Association of Moebius Syndrome (AISMO, www.moebius-italia.it) and by the Regional Health Department as the only national referral centre for the diagnosis and treatment of patients with MBS (e.g., a multidisciplinary approach treating conditions ranging from strabismus correction to smile surgery). This allowed us to contact and follow virtually all of the MBS patients living in Italy. Even MBS patients who currently received medical care elsewhere were evaluated in our hospital at least once in the diagnostic con rmatory phase. All medical data are regularly updated and preserved in the AISMO database, which acts as electronic medical record of these patients, basing on a speci c agreement between our University Hospital and the Association operating since 1999. Any additional information regarding pregnancy with possible assumption of drugs, type of delivery, details regarding the medical history of MBS relatives, are also recorded in the AISMO register. The database is available only for allowed researchers and it is accessible for medical or scienti c purposes on MBS only. As previously reported, MBS cases included in our analysis satis ed the updated "Bethesda Diagnostic Criteria" without any radiological or clinical evidence of further neurological impairment. Subjects not ful lling these diagnostic criteria (in particular, those who were positive for genetic alterations of the TUBB3, HOXA or ROBO3 genes) or presenting karyotype alterations were considered Moebius-like and excluded from the analysis. Every MBS case was periodically evaluated by our multidisciplinary team of physicians, which includes an ophthalmologist (A.C.), a neonatologist or paediatrician, a speech therapist, an orthodontist, an orthopaedist (for children with clubfoot or nger anomalies), and a maxillofacial surgeon with expertise in smile surgery; the frequency of visits depended on the severity of the disease, usually ranging from 6 to 24 months. Each visit includes a comprehensive ophthalmological evaluation of extraocular motility and refraction under cycloplegia in paediatric cases. A detailed history was obtained at the rst visit from relatives of each patient and updated at every visit. All patients (or relatives if minors) have previously given the AISMO consent to use their demographic data for research purposes and statistical analysis. The present series takes into account also patients who had been encompassed in previous research papers. 1,15 The study adhered to the tenets of the Declaration of Helsinki and was approved by the local ethics committee (No. 93/2019/OSS*/UNIPR/June 14, 2019).

Statistical analysis
The authors used data made available by AISMO to obtain the following information for each registered member: date of birth, date and age of diagnosis, gender, and place of provenance/residence. The subjects were assigned to the ve geographical regions (i.e., Northeast, Northwest, Central, South, and Islands) conventionally used for surveys and epidemiological investigations in Italy. Every study parameter was then calculated by reference to the entire country and each region. The rates and prevalence (number of cases per 100000 people) of MBS were calculated referring to the 2018 Italian census performed by the National Institute of Statistics. 16 The rate and corresponding prevalence were calculated for the entire population and each gender (i.e., affected males/Italian males; affected females/Italian females). Concerning the birth prevalence, it was calculated by referring to the year of birth of each patient. Chi-square (c2) analysis was used to investigate possible differences among geographic regions and gender. For gender tests, unweighted (i.e., each rate was weighted equally across regions, independently of the actual gender population in the corresponding region) and weighted (i.e., each rate was scaled to the corresponding gender weight, according to the population density in each region) data were analyzed. A p-value <0.05 was considered statistically signi cant.
Data Availability Statement: data used for the study can be checked at the following links: www.moebius-italia.it and www.istat.it.

Descriptive statistics
The AISMO register contained 231 subjects. Of these, 59 subjects did not fully satisfy the "Bethesda Criteria or they had genetic pro le not compatible with MBS (i.e. karyotype anomalies, TUBB3, HOXA family or ROBO3 mutations); 4 subjects were not Italian citizens (4 subjects) and 4 subjects had incomplete data. In total 67 subjects (29%) were excluded from our cohort. The remaining 164 MBS patients (73 men, 44.5%) were considered for the study and analyzed for epidemiological purposes (Table 1). All patients were Caucasian and all cases were sporadic. The median age at diagnosis was 3.6 years, ranging from 0-55 (a value of 0 means that the diagnosis was achieved within the rst 12 months of life); this range was signi cantly reduced to 0-5 years (median age at diagnosis: 2.2 years) for patients included after 2007.
The Figure 1 shows the newly recorded cases in the AISMO register based on the year of birth; a progressive increase in recorded cases is evident from 1998, when the AISMO register was established, with the new diagnoses peaking (16 cases) in 2005-2006 (coincidentally after the Consensus Conference on Moebius Syndrome, which produced the "Bethesda Criteria"). The birth prevalence calculated referring to the most recent available national data (2016) was 0.06 cases per 10000 live births. The mortality rate in our cohort is 0 (zero) as there was no case of known death. In our series we recorded two cases of homozygous twins both affected by the disease even if with different clinical expression; we have no case of affected siblings and almost all cases were bilateral (8 cases were bilateral but asymmetric and 4 were monolateral). Extra-ocular associated features of our cases are summarized in Table 2. Poland's syndrome 2%

Statistical analysis
The relative rate of MBS in the Italian population was calculated for males and females in terms of the total number of patients diagnosed and separately for each region ( Table 1). The rates were evenly distributed across the different regions, with the exception of the Northwest, where more cases were located ( Figure 2). Moreover, in this region, the gender rate differed since there were 1.5-times more females than males (62.7% vs. 37.3%). The difference in the rate in the Northwest was con rmed with reference to both the total population (P-MT: 0.12 vs. P-FT: 0.20) and the genderbased subdivisions (P-MM: 0.24 vs. P-FF: 0.39), see Table 3. The MBS rate of the total population (i.e., both males and females) subdivided into the ve regions differed signi cantly (p < 0.001; c 2 -test). To verify whether this result was due to the larger population in the Northwest, the data were weighted to account for the different populations in the ve regions. This analysis involved weighted rate data (i.e., data rescaled to the actual population in each region). After this adjustment, no statistical signi cance was observed. Finally, c 2 -test for the separate sex rates was performed. The analysis of males con rmed that the distribution across different regions was similar (no statistical differences for both "unweighted" and "weighted" data over the male population in the different regions). The analysis of females showed a signi cant (p < 0.001) difference for "unweighted" data. However, this difference lost signi cance when "weighted" data over the female population in different regions were considered. An analysis of the total weighted population dataset showed no signi cant differences comparing the ve regions.

Discussion
This study examined the epidemiology of Moebius syndrome in a well-de ned population over a precise geographical area, using de nite diagnostic criteria for MBS. In our series, this disease affects males and females equally, supporting the evidence that MBS is not an inherited X-or Y-related disease. The non-hereditary nature of this disease is supported by the fact that all cases were sporadic in our series as in most reported series. This observation is in line with the recent hypothesis that MBS has a multifactorial basis and genetic mechanisms play a minor role. [7][8][9]17 We found that the overall prevalence of MBS was 0.27/100000 newborns. This value can be reasonably rounded to 0.3/100000 (0.0003%) because few cases of MBS may have missed the appointment with our medical staff and/or not having been enrolled in the AISMO register. This may occur when dealing with people/families affected by congenital genetic disorders, which may carry social stigma leading to self-marginalization. Regardless, the prevalence observed in our study is different from data reported in different parts of the world. For example, in a Dutch series, Verzijl et al. in 1996 estimated a prevalence of 0.002% (i.e. about ten times higher than our results) for MBS, but with poor details on their sources and with different diagnostic criteria with respect to those presently adopted. 13 Similar higher prevalence was reported by physicians with expertise on MBS in the United States, Sweden, and Brazil, but without any populationbased analysis. [10][11][12] Based on our data, we can con rm that MBS is an extremely rare disease. Furthermore, our data may be considered if a dedicated national healthcare program is planned or re-organized. Another important consideration is that we found a uniform distribution of MBS cases in the ve areas considered. These ve regions were conceived by ISTAT, as people living in them have different social, economic, and working lifestyles, with different climates characterizing each region. As we did not identify a region with a higher prevalence of MBS cases, we can exclude environmental factors such as pollution, weather conditions such as intense cold or heat, and prolonged sun exposure during pregnancy as causative for MBS. It appears that the environment had little or no in uence on the disease pathogenesis in our population. The only reported agent that signi cantly increased the risk of newborns being affected by MBS (by a factor of 30) is the use of misoprostol during the rst trimester of pregnancy. 18 Misoprostol (a synthetic prostaglandin E analogue) is an illegal abortifacient widely used in Brazil and other countries in South and Central America. As misoprostol is not in use in Italy, our epidemiological data on MBS lack any pharmacological bias, at least as far as misoprostol is concerned. Beyond misoprostol, in the electronic medical chart of each patient, made available by the AISMO register, we could also exclude the assumption of vasoconstrictor agents, cocaine, abdominal trauma during pregnancy of all the considered patients. In our series all the cases were sporadic and no affected siblings were present. Otherwise, we support the genetic counseling of the relatives with affected offspring as it is postulated that there is a 30-50% risk of recurrence in the familiar trees. Genetics may be also valuable for a better understanding of the pathogenesis of those forms presenting extra-ocular associated anomalies. 19

Conclusion
Our data add elements to the knowledge of MBS providing its exact epidemiology in a highly populated European Country, which may be particularly useful when devising medical policies regarding this rare disease. Most rare diseases are considered "orphans" with no effective treatment; people affected are more vulnerable psychologically, socially, economically, and culturally, as they usually have no response for their medical condition. These di culties can be overcome and the efforts made by the scienti c community can increase our knowledge and give new hope for future treatments of this disorder.
The English in this document has been checked by at least two professional editors, both native speakers of English.  The prevalence of MBS (no. of cases/100,000 people) was determined for each geographic area (Table 2), with an overall prevalence of 0.27/100,000.