Identification of novel genetic, neurobiological and radio-anatomical biomarkers for risk stratification of sudden unexpected death in infancy and early childhood: the BIOMINRISK study protocol

Abstract

Introduction

The BIOMINRISK project is a national French study aimed at identifying novel biomarkers associated with sudden unexpected death in infancy (SUDI) through a multidisciplinary approach encompassing three key components of intrinsic vulnerability to SUDI: genetic, neurobiological and radio-anatomical. A better understanding of the pathophysiological mechanisms underlying SUDI may enhance the personalisation of prevention strategies and contribute to reducing its incidence.

Methods and analysis

We will analyse data from 250 children under the age of 2 included in the national SUDI registry (the OMIN registry) since 2020 for which biological samples and medical imaging data will have been collected from 15 participating French hospitals. Our investigations will focus on three axes: (1) genetic: we will conduct whole genome sequencing family trio analyses to identify novel variants and genes associated with sudden infant death syndrome (SIDS) by examining SIDS cases along with their two parents; (2) neurobiological: a case-control study will be performed to investigate the roles of various neuromodulators—including serum serotonin, blood butyrylcholinesterase and cerebrospinal fluid orexin—in the arousal regulation in children who have died from SUDI. We will recruit 250 living age-matched and sex-matched controls who will undergo blood tests and lumbar punctures as part of their routine care and (3) radio-anatomical: a case-control study will explore the potential anatomical predisposition to SUDI by assessing upper airway narrowness. We will compare the osseous structures of the upper airways (nasal fossae, hard palate) using geometric morphometrics on CT images. Recruitment of 250 living age-matched and sex-matched controls who have undergone brain CT scans, including facial bones, will be conducted.

Ethics and dissemination

The study has received ethics approval for all three axes. Results will be published in international peer-reviewed journals and presented at national and international conferences.

Trial registration number

NCT06244433.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The collaboration with the national registry for sudden unexpected death in infancy (SUDI), the OMIN registry, ensures comprehensive enrolment across France, providing high-quality, standardised clinical data and biological samples from SUDI cases.

• The genetic analysis uses a trio approach for whole genome sequencing (including the deceased infant and both parents), enabling the identification of ultra-rare de novo pathogenic variants and significantly enhancing the diagnostic probability compared with genomic analysis of the child alone.

• The neurobiological and radio-anatomical analyses employ a case-control design, allowing for the simultaneous investigation of multiple SUDI risk factors while controlling for potential confounding variables through comparison with age-matched and sex-matched living controls.

• Some research hypotheses are exploratory, which may result in an insufficient number of included children to ensure statistical power for certain analyses.

Introduction

Background and rationale

Sudden unexpected death in infancy (SUDI), defined as the unexpected death of an apparently healthy infant without obvious cause before investigation, remains one of the leading circumstances of post-neonatal infant mortality in Europe and other high-income countries, including the USA, with notable disparities among countries.^{1 2} Over the past three decades, epidemiological studies have identified significant environmental risk factors, such as prone sleeping position or maternal smoking.^{3 4} In the 1990s, prevention campaigns targeting these risk factors, particularly promoting safe sleeping positions, led to a substantial decline in SUDI incidence. However, since the early 2000s, SUDI rates have stagnated in France and many other countries.^{1 5 6} This stagnation may be attributed to the complex pathophysiology of SUDI, which remains incompletely understood due to the diversity and interrelatedness of the contributing factors and mechanisms. Each case of SUDI is likely the result of multiple interacting factors, complicating the identification of a single cause or universal sequence leading to death. Thus, despite complete postmortem investigations, 50%–70% of SUDI cases remain unexplained and are subsequently classified as sudden infant death syndrome (SIDS).^{5 7 8} The prevailing hypothesis regarding the mechanisms involved in SUDI is the ‘triple risk’ model, which suggests that death occurs due to the coexistence of environmental risk factors and intrinsic vulnerabilities in infants during a critical developmental period.⁹

Recent expert recommendations have emphasised the importance of investigating the intrinsic vulnerabilities of infants, focusing on genetic, biological and anatomical factors related to dysfunctions in the cardiac, respiratory and/or arousal systems.¹⁰ Genetic variants with functional effects in genes associated with channelopathies, cardiomyopathies or metabolic diseases have been identified in 10%–20% of SIDS cases.¹¹,¹⁷ While some of these mutations may directly cause death, others may contribute to mortality when combined with environmental risk factors.¹⁸ Research has also examined various neurobiological systems and pathways involved in arousal regulation, highlighting the brainstem’s crucial coordinating role. Dysfunction in these systems may hinder normal protective responses to environmental stressors during sleep.^{19 20} Among the studied systems, abnormalities in serotonergic, cholinergic and orexinergic pathways appear to be particularly significant in cases of SIDS.²¹,²⁶ Furthermore, anatomical variations in the upper airways (UA) may serve as precipitating factors for asphyxia-related deaths.^{27 28} Specifically, anatomical variants such as the ogival (high and narrow) palate have been clinically and radiographically documented in children who have died of SUDI, as well as in those with obstructive sleep apnoea syndrome.²⁹,³³

Despite the identification of environmental and intrinsic vulnerability risk factors, their presence is not consistently observed, resulting in numerous unanswered questions for families and hindering the implementation of targeted preventive measures. This challenge has prompted our exploration of new genes, biological pathways and anatomical variations as underlying factors contributing to an infant’s intrinsic vulnerability to SUDI.

The overall aim of the BIOMINRISK project is to concurrently investigate three key intrinsic vulnerability components—genetic, neurobiological and radio-anatomical—to identify novel biomarkers associated with SUDI. This multidisciplinary approach addresses the heterogeneous pathogenesis of SUDI and may facilitate the development of personalised prevention strategies, benefiting both the general infant population and families affected by a case of SUDI. Ultimately, this effort aims to reduce avoidable deaths.

Objectives

The BIOMINRISK project is organised into three axes—genetic, neurobiological and radio-anatomical—each corresponding to a distinct clinical study with specific objectives. All the axes focus on the same population of children who have died from SUDI, with a particular emphasis on SIDS. Depending on the axis, parents or living controls will also be included. In France, children who die suddenly before the age of 2 are classified as SUDI, even though the pathophysiology of sudden death may differ before and after the age of 1.³⁴ This classification enables standardised care for the families of these young children and allows for the collection of clinical and epidemiological data in the French national registry for SUDI up to the age of 2. For the sake of clarity, we will also use the term ‘SUDI’ to refer to children between the ages of 1 and 2, even though this group more accurately falls under the definition of sudden unexpected death in childhood.³⁵

For the genetic axis, the primary objective is to identify novel genes and variants associated with SIDS through whole genome sequencing of family trios, which include the deceased infant and both biological parents. The secondary objectives are: (1) to identify new compound heterozygous variants and/or copy number variants implicated in SIDS and (2) to evaluate genotype-phenotype associations and correlations in SIDS by cross-analysing clinical and biological characteristics studied in the neurobiological and radio-anatomical axes.

For the neurobiological axis, the overall aim is to understand the role of various neuromodulators in SUDI by simultaneously exploring three arousal systems: serotonergic, cholinergic and orexinergic. For the serotonergic system, we will compare serum serotonin concentrations between cases of SIDS and a population with explained SUDI. For the cholinergic system, we aim to compare blood butyrylcholinesterase (BChE) activity between cases of SUDI and living controls, accounting for the timing of sample collection. For the orexinergic system, we will: (1) compare cerebrospinal fluid (CSF) orexin concentrations in infants who died from SUDI and living controls and (2) assess the association/correlation between specific medical, sociodemographic or environmental factors and CSF orexin concentrations in infants who died from SUDI.

For the radio-anatomical axis, the objective is to investigate differences in the anatomical conformation of facial and cranial bone structures—including the midface. (nasal fossa, maxilla), mandible and skull—between infants who died from SUDI and living controls, using CT scans. To complement this comparative analysis, we will create a mapping of the UA of living children to establish normative anatomical data for the bony structures of the UA in infants under 2 years of age, providing metric and morphological information categorised by age and sex.

Method and analysis

General study design and settings

The BIOMINRISK project will be conducted across 15 participating centres in France, including the University Hospitals of Amiens, Angers, Besançon, Bondy, Brest, Clamart, Grenoble, Lyon, Marseille, Montpellier, Nancy, Nantes, Rouen, Saint Etienne and Toulouse. Each of the three axes corresponds to a multicentre clinical study where SUDI cases and controls will be enrolled either retrospectively (using existing data since 2020) or prospectively. Recruitment started on 13 July 2024, with an anticipated inclusion period of 24 months, projecting an end date of July 2026. The study designs and methods for each axis are outlined below. The study protocol has been developed following the ‘Standardised Protocol Items: Recommendations for Observational Studies’ guidelines and checklist (see online supplemental appendix 1).³⁶ The research will be conducted in collaboration with the French national registry for SUDI, known as OMIN (Observatoire national de la Mort Inattendue du Nourrisson; https://www.omin.fr/).^{37 38} This registry has been documenting all SUDI cases in children under 2 years of age in France since 2015, collecting prospective demographic, clinical, biological, imaging and autopsy data via an electronic case report form (eCRF) completed by referral paediatricians in the 37 referral centres participating in the registry. Since 2020, a biobank (BioMIN, No. DC-2019-3850) linked to the registry has been established, containing various biological samples (blood, urine, CSF, hair and faeces) from SUDI cases.

Study population: SUDI cases

The three axes of the BIOMINRISK project will include the same cohort of children who have died from SUDI, aged from birth to 2 years. The general inclusion criteria are as follows: children must be enrolled in the national OMIN registry after obtaining written informed consent from their parents; blood and CSF samples must be available in the BioMIN biobank; parents must consent to access the birth blotter generated as part of the national systematic neonatal screening programme; and the children must have undergone postmortem CT scans of the head and/or whole body. SUDI cases will not be included if they exhibit any of the following conditions: known metabolic, genetic or syndromic diseases at the time of death; evidence of neuromeningeal viral, bacterial or parasitic infections in the CSF or unusable CT scan images due to artefacts, incomplete acquisition or reconstruction (table 1).

Table 1. Inclusion and non-inclusion criteria.

Inclusion criteria	Non-inclusion criteria
For SUDI cases
Aged from birth to under 2 years. Deceased in the context of SUDI. Included in the national OMIN registry with written informed parental consent. Blood and CSF samples available in the BioMIN biobank. Access to the birth blotter from the national systematic neonatal screening programme. Postmortem CT scan of the head or whole body conducted.	Known metabolic, genetic or syndromic diseases at the time of death. Evidence of neuromeningeal viral, bacterial or parasitic infections in the CSF. CT scan images that are unusable due to artefacts, incomplete acquisition or reconstruction issues.
For parents of SUDI cases (genetic axis)
Biological parent of a SUDI case enrolled in the BIOMINRISK project. Written consent to participate in the BioMIN biobank. Affiliation with an appropriate health insurance system.	Known metabolic, genetic or syndromic diseases. Individuals under guardianship or curatorship.
For living controls (neurobiological axis)
Aged from birth to under 2 years. Blood sample and lumbar puncture performed as part of routine care. Access to the birth blotter from the national systematic neonatal screening programme. Parental non-opposition to participation. Parental affiliation with an appropriate health insurance system.	Known metabolic, genetic or syndromic diseases. Evidence of neuromeningeal viral, bacterial or parasitic infections in the CSF.
For living controls (radio-anatomical axis)
Aged from birth to under 2 years. CT scan of the head and/or neck, including facial bones. Parental non-opposition to participation. Parental affiliation with an appropriate health insurance system.	Known metabolic, genetic or syndromic diseases. Alterations in normal anatomical relationships (eg, major syndromic dysmorphia, tumours, facial trauma). CT scan conducted due to severe malaise. Death occurring before the age of two in the context of SUDI. CT scan images deemed unusable.

CSF, cerebrospinal fluid; SUDI, sudden unexpected death in infancy.

SUDI cases will be classified using two criteria: age (under or over 1 year) and cause of death based on international recommendations.³⁹ Cases will be categorised either as explained SUDI, when a cause of death is identified through postmortem investigations, or as unexplained SUDI, when no cause can be determined despite thorough complete investigations—in which case the death is classified as SIDS. We anticipate recruiting a total of 250 SUDI cases from the national OMIN registry during the recruitment period. As of July 2024, 85 eligible SUDI cases have already been identified within the OMIN registry for inclusion in the three axes of the BIOMINRISK project.

Genetic axis

Study design

We will conduct a national, observational, multicentre genetic study linked to a biobank of blood samples from parental couples for trio analyses. Whole genome sequencing will be performed on children who died from SIDS and their biological parents, using a high-throughput method. This approach allows us to examine all coding and non-coding sequences in the genome without prior hypotheses, helping us identify pathogenic allelic variants. The data will be analysed using a trio approach to find de novo variants—those found in the infant but not in either parent’s genome—focusing on ultra-rare pathogenic variants (minor allele frequency <0.01%).

Specific population: parents of SUDI cases

We aim to include 150 couples (300 parents) from the 250 SUDI cases prospectively. The inclusion and exclusion criteria for parents are detailed in table 1. We expect about 95 of these cases will be classified as SIDS, involving 190 parents.

Outcome measures

The primary outcome is the identification of de novo genetic point mutations in coding and non-coding sequences among the SIDS cases included in the trio analyses. The secondary outcome focuses on finding compound heterozygous variants and/or copy number variants in the SIDS cases.

Neurobiological axis

Study design

This national, multicentre, observational case-control study will analyse 250 SUDI cases from the BIOMINRISK project and 250 living matched controls (see below). The study focuses on three neuromodulators involved in arousal regulation: serum serotonin, blood BChE activity and CSF orexin. Samples for SUDI cases will be sourced from the BioMIN biobank, while blood and CSF samples for living controls will come from routine clinical care. No additional procedures will be conducted on the living controls.

Specific population: living matched controls

The 250 living controls will be infants under 2 years of age who have undergone blood and CSF sampling during routine care and who have no evidence of neuromeningeal viral, bacterial or parasitic infections in the CSF. Controls will be matched to SUDI cases by age and sex. Inclusion and non-inclusion criteria are detailed in table 1.

Outcome measures

Serum serotonin concentration will be measured only in SUDI cases due to known physiological changes postmortem. BChE activity will be assessed in both SUDI cases and living controls using whole blood samples collected on blotting paper. Two time points will be analysed: at birth, as part of the national neonatal screening programme and at death for SUDI cases or during hospital admission for living controls. CSF orexin concentration will be measured in both SUDI cases and living controls.

Radio-anatomical axis

Study design

This national, multicentre, observational case-control study will analyse 250 SUDI cases from the BIOMINRISK project and 250 living matched controls (see below), distinct from those in the neurobiological axis. CT scans performed during postmortem investigations of SUDI cases and routine clinical management of living controls will be used to examine and compare the anatomical conformation of UA bone structures in these two populations. Anatomical structures will be studied using a geometric morphometric approach, modelling the mean shapes of midface structures.

Specific population: living matched controls

The 250 living controls will be retrospectively selected from children under 2 years of age who have undergone a CT scan of the head and/or neck, including facial bones. Screening will be based on imaging data available in the Picture Archiving and Communication System of the participating centres. Controls will be matched to SUDI cases by age and sex. Inclusion and non-inclusion criteria are detailed in table 1.

Outcome measures

The study will assess the shape of midface structures, including the height and width of the hard palate, nasal cavity volume (nasal cavity width and height, choanal width, nasal cavity asymmetry), the angles of the nasion/turcica sella/maxilla/mandible and nasion/turcica sella/mandible, skull shape (with or without deformity) and cranial perimeter. The radiological impression of an ogival palate on native images will also be evaluated. Measurements and shape comparisons will be conducted using manually and semi-automatically placed landmarks on 2D native images and 3D shape reconstructions, with segmentation software. Affine image registration will be used to create a midface template for both SUDI cases and living controls, generating colourimetric maps to highlight anatomical differences between deceased and living children. The template of living controls will serve as a reference to precisely map the normal distribution of UA structures in children under 2 years of age.

Study timing and data collected

After confirming that the child or parent meets the inclusion criteria for their respective study group, clinical, biological and radiological data will be collected at the inclusion visit (figure 1).

Figure 1. Study schedule of enrolment, examinations and assessments. SIDS, sudden infant death syndrome; SUDI: sudden unexpected death in infancy.

For SUDI cases, data will include demographic characteristics, personal and family medical history, exposure to SUDI risk factors, ante-mortem signs suggestive of obstructive airway disorders, circumstances of death, findings from postmortem investigations and classification of the cause of death. These data will be collected from the eCRF of the national OMIN registry.

For parents included in the genetic axis, age and sex will be documented. Blood samples of 8–10 mL will be collected from both biological parents in EDTA tubes for whole genome sequencing.

For living controls, demographic and clinical data such as age, sex, gestational age and medical history will be collected. In the neurobiological axis, blood and CSF samples will be obtained from surplus volumes collected during routine hospital care at the time of admission. Blood samples on blotting paper for BChE activity measurement at birth, for both SUDI cases and living controls, will be retrieved from the regional centres of the national neonatal screening programme. For the radio-anatomical axis, the clinical indication for imaging and the corresponding CT scan data will be extracted from the child’s medical record. On completion of genetic, biological and radio-anatomical analyses, results will be systematically entered into the eCRF.

Statistical analysis

Genetic axis

Whole genome sequencing data from family trios will be analysed to identify de novo genetic point mutations in both coding and non-coding sequences. A bioinformatics pipeline will be applied, using haplotype-linkage analysis and identity-by-descent methods. The first step involves aligning sequencing reads to the reference genome to determine their genomic location. This data will be stored in text format in a compressed binary alignment map file. Variant annotation will be performed using the GATK HaplotypeCaller algorithm, which will compare sequencing reads with the reference genome to generate a list of single nucleotide variations and insertions/deletions in variant call format. These variants will then be annotated using VEP and SnpEff tools and filtered based on multiple criteria, including variant quality (sequencing coverage, allelic ratio), predicted protein impact, frequency in control populations, bioinformatic pathogenicity predictions and presence in clinical databases, such as ClinVar.^{40 41} For analysis, variants with functional consequences classified as missense, stop_lost, start_lost, frameshift_variant, stop_gain, splice_acceptor_variant or splice_donor_variant will be retained. Only variants with a maximum allelic frequency of 0.1% in the gnomAD database will be considered, reflecting the rarity of the disease. Additionally, variants must have high pathogenicity prediction scores. In the absence of a known family history of sudden death, a de novo variant in the index case will be assumed. To identify novel genotype-phenotype associations and correlations, statistical comparisons between genetic variants and clinical or biological characteristics of SIDS cases will be conducted. Quantitative variables will be analysed using Student’s t-test or Mann-Whitney test, while qualitative variables will be compared using χ² or Fisher’s exact tests.

Neurobiological axis

Serum serotonin concentration, blood BChE activity and CSF orexin concentration will be summarised using mean (SD) and median (IQR) values across all SUDI cases, as well as within the subgroups of SIDS cases and explained SUDI cases and age subgroups, and living controls.

To investigate differences in the serotonergic, cholinergic and orexinergic systems, logistic models will be applied while accounting for potential confounders such as gestational age and sex. These models will be used to: (1) compare serotonin concentrations between SIDS cases and explained SUDI cases; (2) compare blood BChE activity between SUDI cases and living matched controls at multiple time points—(a) at birth and at death in SUDI cases; (b) at birth and at hospital admission in living controls; (c) at birth between SUDI cases and living controls and (d) at the time of death for SUDI cases vs at hospital admission for living controls and (3) compare CSF orexin concentrations between SUDI cases and living matched controls and between SIDS cases and explained SUDI cases.

To explore the relationship between CSF orexin concentrations in SUDI cases and medical, sociodemographic or environmental factors, correlation analyses (Pearson or Spearman tests) will be performed for quantitative variables, and association analyses (Mann-Whitney or Kruskal-Wallis tests) will be used for qualitative variables. These analyses will also be conducted separately for the subgroups of SIDS cases and explained SUDI cases.

Radio-anatomical axis

Statistical analyses will be performed on each landmark and group of landmarks to assess anatomical differences between SUDI cases and living controls. Generalised Procrustes analysis will be applied to normalise and rescale coordinate data for shape comparisons. Principal component analysis will visualise the distribution of anatomical variations within each group. Goodall’s F Test, combined with canonical variates analysis, will identify significant morphological differences between SUDI cases and living controls. Centroid size comparisons will be conducted using either the Wilcoxon test or paired t-test. Subgroup analyses will be performed on age, sex, cranial anomalies, subjective assessment of an ogival palate, causes of death, ante-mortem clinical signs and weight trends from birth to death in SUDI cases.

To establish a comprehensive reference for normal bony anatomy of the UA, a detailed descriptive analysis of the 250 living controls will be performed, including calculation of means, medians, SD and quartiles. A template of facial and cranial bone structures will be generated to serve as a reference model for anatomical conformation in children under 2 years of age.

Statistical analyses will be conducted using R statistical software V.4.4.2 (R Project for Statistical Computing), incorporating specialised packages such as shapes, geomorph and mshapes, procGPA, procD.lm, procdist and CVA to ensure precise morphological comparisons.

Sample size justification

The number of SUDI cases included in the study is based on the recruitment capacity of the national OMIN registry during the study period. Given these constraints, statistical power calculations will be performed a posteriori to assess the robustness of the findings.

Patient and public involvement

The BIOMINRISK project is supported by three associations representing parents affected by SUDI: Naître et Vivre (https://naitre-et-vivre.org/), Sa Vie (https://www.sa-vie.fr/), and Les Rires d’Anna (https://www.helloasso.com/associations/les-rires-d-anna). Through our collaboration with these organisations, we have developed thoughtful strategies for engaging with parents participating in the genetic aspect of the study, ensuring our communication is both respectful and sensitive to their experiences of loss. The results of the BIOMINRISK project will be disseminated to the general public through these associations.

Ethics and dissemination

Ethics approval and consent to participate

The protocols for the three study axes have received approval from the ethics committee in France: the genetic axis was approved in December 2023 (No. 2023-A01993-42), the neurobiological axis in October 2023 (No. 23-128-10-190) and the radio-anatomical axis in September 2023 (No. 23-116-09-212). Children who die in the context of SUDI are routinely included in the national OMIN registry and the BioMIN biobank after obtaining informed and written consent from their parents. The data collected in the national OMIN registry comply with the French Data Protection Authority (Commission Nationale de l'Informatique et des Libertés), which granted approval for the registry in 2015 (authorisation No. 915273). No additional procedures are necessary for the inclusion and examinations of SUDI cases, as the samples and imaging tests used in this project are standard practices conducted to investigate SUDI, following French guidelines. Specific written consent for the genetic study will be systematically obtained from participating parents prior to their inclusion (the consent forms can be provided by mail at omin@chu-nantes.fr). Parents of children enrolled as living controls in the neurobiological and radio-anatomical axes will receive a written document outlining the study, and they will have the opportunity to decline their child’s participation within 1 month, in accordance with French legislation.

Data collection, storage, monitoring and confidentiality

Data collection for each child and parent participating in the research will be conducted using an eCRF. For SUDI cases, relevant data will be pre-existing in the eCRF of the national OMIN registry. Biological samples from all participants will be anonymised with a predefined code before being sent to designated facilities: l’institut du thorax in Nantes (INSERM UMR 1087) for genome sequencing via its genomics core facility (GenoA) and bioinformatics core facility (BiRD), the Biochemistry and Molecular Genetics Department of the University Hospital of Clermont-Ferrand for BChE activity analyses, and the Lyon Neuroscience Research Centre (INSERM U1028/CNRS UMR 5292) for serotonin and orexin analyses, all under a collaborative agreement.

Each individual responsible for entering data into the eCRF (investigator or research staff) will have a unique user account with specific access rights based on their role, allowing them to enter, modify, lock, monitor or sign eCRF pages. Data will be compiled directly from the eCRF into a secure database hosted on a server with controlled access (account/password) according to user roles. All changes to the data—whether additions, modifications or deletions—will be recorded in a non-editable electronic file (audit trail). Anonymised CT scan images will be uploaded by investigators and securely stored on a web central database for the duration of the study. All investigators will retain study data for a minimum of 15 years after the study’s conclusion. To maintain confidentiality, each participant will be assigned a code that will be used in the eCRF and all related documents (laboratory samples, CT scans, etc), serving as the sole means of linking data back to individual participants.

Monitoring of data within the national OMIN registry will follow standard procedures to ensure data quality. If necessary, adjustments and contingency plans will be implemented to address issues such as under-inclusion or over-inclusion, ensuring that the sample aligns with the predetermined composition and number of participants across the various study axes.

Dissemination

For widespread dissemination, the results of this project will be presented at national and international congresses, published in international peer-reviewed journals, and communicated to the general public through associations of parents affected by SUDI.

Discussion

Given the existing gaps in our understanding of the pathophysiology of SUDI, an innovative multidisciplinary approach such as proposed in the BIOMINRISK project may effectively address the heterogeneous nature of this condition. This French nationwide project aims to explore novel biomarkers for SUDI, ultimately developing personalised preventive measures based on these biomarkers. This initiative seeks to benefit both the general population of infants and families affected by SUDI, contributing to the reduction of avoidable deaths.

A significant strength of this study lies in its simultaneous exploration of three key components of intrinsic vulnerability to SUDI—genetic, neurobiological and radio-anatomical— within a large cohort of children who have died from SUDI. Each research axis is constructed around innovative methodologies developed from the current state of knowledge in the field. The anticipated results could have multiple implications. First, gaining a deeper understanding of the complex mechanisms underlying SUDI, identifying new pathways and their potential interactions, may open avenues for further research, particularly in pharmacology, to address possible dysfunctions in the serotonergic, cholinergic and orexinergic systems. Second, the identification of predictive biomarkers could enhance existing SUDI risk prediction algorithms, better accounting for intrinsic vulnerability factors when assessing an infant’s risk of SUDI. This would facilitate the development of personalized preventive strategies.⁴² For instance, if anatomical susceptibility to SUDI is confirmed—such as a narrower UA—antenatal ultrasound screening followed by clinical assessments at birth, including palatal examinations, could be implemented.²⁸ In the most vulnerable infants, enhanced preventive measures related to sleep, such as recommending pacifier use—known to be a protective factor against SUDI—could be reinforced.⁴³

From a methodological standpoint, this study possesses several strengths. The national OMIN registry, the only comprehensive registry for SUDI in Europe, will be instrumental in achieving the target number of cases while ensuring exhaustive enrolment across France.^{37 38} This registry will also provide high-quality standardized clinical data on SUDI cases, complemented by consistent biological samples from the associated biobank. The case-control design employed for the neurobiological and radio-anatomical axes will facilitate the simultaneous examination of multiple SUDI risk factors while controlling for potential confounding variables through comparisons with age-matched and sex-matched living controls. In the genetic axis, trio analysis presents a robust approach for identifying ultra-rare pathogenic variants that arise de novo, significantly enhancing diagnostic capabilities compared to standard genome or exome analyses of the child alone.⁴⁴,⁴⁶ This method can help distinguish genetic variants likely associated with rare diseases from those that are not, as many variants can be excluded through comparative analysis with parental sequences.

Nevertheless, the study has limitations. The final enrolment numbers will depend on the actual incidence of SUDI during the project, which remains unpredictable. As some of the hypotheses being tested are exploratory and innovative, the number of children included may not provide adequate statistical power for certain analyses, potentially leading to less precise results. Additionally, the retrospective inclusion of some participants may introduce risks of missing data and information bias.