Introduction

In severely ill children hospitalized in the neonatal intensive care unit (NICU) or pediatric intensive care unit (PICU), a quick etiological diagnosis is key to proper care management. The phenotypic and genetic variability of severe early genetic conditions and the lack of developmental data make a phenotype-first approach challenging. Exome (ES) and genome sequencing (GS) have been used in research studies for decision-making and care of critically ill patients with a proof-of-principal in 2012 [1] then in larger studies when the turn-around time from blood sampling to results became compatible with the time constraints of the NICU/PICU [2,3,4,5,6].

In France, current genetic testing with no phenotypic diagnosis includes as first-tier testing array-comparative genomic hybridization array (CGH-array). Further testing is carried out using gene panels, ES or GS. This lengthy process does not suit the timeline of critical care. The aim of this study was to evaluate the possibility of on-site rES in a French regional university hospital of getting a rapid genetic diagnosis in critically ill infants.

Methods

Patients

The inclusion criteria were: children under one year admitted to intensive care who would benefit from an early diagnosis as evaluated by the NCIU/PCIU team, suspicion by a consultant of the clinical genetics team of a monogenic condition (i.e.,: multi-systemic involvement, severity of symptoms, consanguinity), both biological parents available (Fig. 1). Patients were excluded if a genetic diagnosis was clinically recognizable, if a chromosomal anomaly was thought more likely or if another non-genetic condition could explain the phenotype.

Fig. 1: Flowchart of included individuals.
figure 1

BWS Beckwith Wiedemann syndrome, CMV Cytomegalovirus.

Full size image

Study design and ethics

This prospective single-center study recruited patients from April 2019 to April 2021 who were admitted in the NICU or PICU of the University hospital of Montpellier. Informed signed consent was obtained from both parents. This study was accepted by the ethical committee of Montpellier university hospital (CNRIPH 18.07.17.33927).

Outcomes

The primary outcome was the number of results reported to parents within 16 days from blood sampling. Secondary outcomes were the number of diagnosis, the number of cases where rES led to a change in medical care, as defined by a treatment, test, action performed because of the rES results.

Clinical data and concomitant genetic testing

Clinical data was collected from medical records and phenotype was described using the human phenotype ontology (HPO) [7] terms and detailed in Table 1. Patients were clinically reevaluated at 3 months and a year after the inclusion. Other genetic tests, including array-CGH if indicated, were performed as part of standard genetic testing.

Table 1 Phenotype of included individuals.
Full size table

ES processing and interpretation

DNA was extracted from a venous blood sample. We used the SureSelect QTX Library prep (Agilent technologies) enrichment kit and SureSelect Human All Exons V7 (Agilent technologies) following the manufacturer’s instruction. Sequencing was performed on a NextSeq 500 (Illumina). Bioinformatics were performed following Broad institute GATK best practice guidelines [8]. Pipeline for variant calling and prioritization is reported in literature [9].

rES data was jointly interpreted by molecular biologists and clinical geneticists in a multi-disciplinary approach. Classification of variants followed the recommendations of the American College of Medical Genetics and Genomics (ACMG) [10]. The report was written by the molecular biologist. Variants were confirmed with Sanger sequencing and a new report issued.

Results

Population

Fifteen patients were included in this study. The median age at blood sampling was 38 days (extremes: 4–272; interquartile range: 8–82). Phenotypes of patients are summarized in Table 1; results and impact on care in Table 2. Prenatal invasive genetic testing was carried out in 5 mothers of included infants.

Table 2 Results of included individuals.
Full size table

Primary outcome

Results were reported to parents within 16 days from blood sampling in 13/15 cases (86%), with a median of 14 days (extremes 9–145; interquartile range: 13–15; Table 2). All rES results were available within 16 days. For individual 4, the parents declined the scheduled genetic consultation, because of their child’s death. For individual 7, the parents were not available for a consultation before day 16. Of the 15 children, two died before the results.

Secondary outcome

Number of diagnoses after rES analysis

A diagnosis was made by rES in 6 individuals (40%, Table 2). Concomitant tests are summarized in Table 1. In individual 13 diagnosis was made by array-CGH (concomitant genetic test) who had a diagnosis of complex chromosome 18 rearrangement. A variant of unknown significance (VUS) was found in LAMA1 (Laminin alpha-1, MIM 150320) is in trans of an interstitial deletion due to the complex chromosomal rearrangement which could partly explain the phenotype [11]. Individual 5 had a diagnosis of Beckwith Wiedemann syndrome with mosaic paternal uniparental disomy. In individual 11, because of the severity of the phenotype, a double diagnosis (genetic and a cytomegalovirus infection) was suspected. rES overturned this hypothesis.

rES impact in patient’s care

The diagnosis changed patients’ care in patients where rES was positive in four individuals. In individual 1 with a pathogenic variant in the potassium voltage-gated channel subfamily Q member 2, (KCNQ2, MIM 602235), antiepileptic drugs were switched to carbamazepine, rapidly stopping the seizures. For individual 2, as rES excluded a severe metabolic disorder, the ventricular septal defect was repaired. Individual 3 who had a myotubularin gene variant (MTM1, MIM 300415) had his muscular biopsy cancelled and was evaluated for a clinical trial of gene therapy (Clinical trial “ASPIRO”, NCT03199469). In individual 15, the WT1 variant (WT1 transcription factor, MIM 607102) explained the terminal renal failure. Because of the increased tumor risk, she underwent bilateral gonadectomy and nephrectomy with a diagnosis of nephroblastoma at pathological examination. Caryotype further revealed 46, XY chromosomes.

Negative results were also useful in three individuals. For individual 5, rES eliminated differential diagnoses so active care was continued as Beckwith Wiedemann syndrome generally has a good prognosis when properly managed [12]. For individuals 9 and 10, the negative results helped the pediatric team to offer invasive surgical procedures that would not have been considered had there been a severe genetic disease diagnosed by rES.

In individual 8 a VUS was found in ATP7A (ATPase Copper Transporting Alpha, MIM 300011), the gene responsible for Menkes syndrome. After review by the clinical genetics team and the molecular biologists, this variant was reported, because further biochemical testing was necessary with a potential treatment [13].

A VUS was also found in ABCD1 (ATP Binding Cassette Subfamily D Member 1, MIM 300371, c.896A > G;p.His299Arg) responsible for adrenoleukodystrophy [14] in individual 15. Very long fatty chain screening showed slightly increased C26 at 1.38 µmol/L (normal 0.3–1.2). Familial segregation showed that two maternal great-uncles carry the variant with normal very long fatty acid chain screening re-classifying this variant as likely benign, reassuring for her (male karyotype), and family counselling.

Discussion

This monocentric study shows the experience of a French regional university hospital, very close to what can be implemented in clinical routine care. rES was usefull in early diagnosis and medical care. In this study, 6 individuals (40%) received a diagnosis with rES. This yield is comparable to larger studies who have found a variable diagnostic yield (21–72% [15]). Some symptoms have remained unexplained: macroglossia in individual 1 and seizures in individual 7. Seizures have occasionally been reported in individuals with cardio-facio-cutaneous syndrome which is a RASopathy [16] but is not reported as a usual feature of Noonan syndrome including in individuals with LZTR1 pathogenic variants. A negative exome is usefull while discussing the possibility of palliative care: a poor genetic prognosis can direct towards palliative care and negative rES to continue active invasive care. The usefulness of a negative result has also been previously reported [6]. A positive result limits the number of diagnostic invasive tests and cost [17, 18].

Teams considering this strategy as a first-tier genetic test in critically ill children should implement a protocol for dealing with incidental findings of the ACMG list [19], heterozygosity for a genetic disorder with a high rate of heterozygosity in the general population, and VUS. In the case of VUS, further tests can be needed to specify the pathogenicity of the variant. VUS in individuals 8 and 15 highlights the difficulties arising with the uncertainty of the clinical consequences of VUS, the medical and familial stress generated by the subsequent delay of the additional tests. In the diagnostic clinical setting, teams should have a protocol for reporting VUS and incidental findings [19].

Limits to this study include its small size sample. Other studies have reported quicker turnaround time with the shortest of 3.3 days [6]. In their study, 86% (ultra-rES) and 78% (rES) of parents had the result before their child was discharged or died.

The rapid evolution of genetic sequencing technologies as well as the acceptability of costs involved by health services might be hurdles to a quick introduction of the latest sequencing technologies in routine patient care [6]. Concerning funding of rapid pan-genomic sequencing, rES reduces unnecessary, costly, and sometimes invasive procedures as well as the time to diagnosis. It also allows genetic counselling as well as proper care management of the child according to their diagnosed condition [18]. Parental acceptance is good with little adverse reactions [20]. Hence, the medical, psychological, and monetary benefits of rES outweigh the risks.

In conclusion, rES is feasible in critically ill infants in a French regional university hospital with a good diagnostic yield. It can be offered in individuals for whom a genetic condition is suspected where a diagnosis could help in their care. New studies show a possible switch to rGS to diagnose monogenic diseases but also chromosomal abnormalities. Sequencing technology and costs as well as evolution of data storage will make GS more accessible and may soon become the first-tier genetic test in critically ill children with suspected genetic disorders.