INTRODUCTION

Dysmorphism associated with develop­mental delay or mental retardation is a common indication for chromosomal analysis. Although the underlying aetiol­ogy may only be recognized in ~40 to 60% of cases, there is a need to obtain an accurate diagnosis as this will affect overall management in terms of coun­seling for recurrence and prognosis (Manning & Hudgins 2007). Recent studies using array comparative genomic hybri­dization (aCGH) have shown that dele­tions and duplications are causative in 8 to 17% of such cases (Manning & Hudgins 2007). We report an infant with craniofa­cial dysmorphism and global develop­mental delay who was diagnosed to have a terminal chromosomal deletion follow­ing aCGH.

CASE REPORT

A 3 month-old infant was referred for fa­cial dysmorphism. The antenatal period was uneventful and she was born term via normal delivery. She was small for gestational age with birth weight 2.17kg. Her parents were non-consanguinous. She was admitted for neonatal jaundice at day 6 of life; newborn screening for hearing via otoacoustic emission (OAE) was abnormal. Follow-up with brainstem evoked response (BSER) revealed bila­teral sensorineural hearing loss. Clini­cally she was microcephalic and had fail­ure to thrive. She had fair skin pigmenta­tion. The ears were simple; there was hypertelorism and she had a high nasal bridge. Her nose was short; the lips were thin and the corners of her mouth were downturned (Figure 1). Her fingers were clenched and she had bila­teral clinodac­tyly. There was also feeding difficulty i.e. she had swallowing incoor­dination and there was global develop­mental delay. MRI brainshowed micro­cephaly with evidence of white matter volume loss and cerebral atrophy sug­gestive of perinatal ischaemic insult. Screening for inborn errors of metabol­ism was negative and the preliminary conventional karyotyping (of 4 meta­phase spreads) was reported as normal. In view of the craniofacial dysmorphism and global developmental delay, blood was sent for aCGH. The in­vestigation revealed a 30.6Mb deletion at 5p15.33-13.3 (Figure 2) which was con­firmed by fluorescence-in-situ hybridiza­tion (FISH) using the Vysis LSI EGR1/D5S23, D5S721 Dual Color Probe as shown in Figure 3.

DISCUSSION

Chromosomal analysis is often re­quested as a first-line investigation in a child with mental retardation, multiple congenital malformation and dysmor­phism. Based on data from published research, mental retardation affects ap­proximately 3% of the general popula­tion; of these, less than 50% have an identifiable cause of which 40% is attri­butable to chromo­somal anomaly (Manning & Hudgins 2007).  In a review by Sharkey et al. 2005, 0.8% of liveborn in­fants have chromosomal ano­malies of whom half are associated with an ab­normal phenotype (Sharkey et al. 2005). With conventional moderate level karyo­typing (at 400 to 500 banding), the detection (rate of chromosomal anomalies is 13.3%; 3.8% was for structural and 7.8% for nu­merical anomalies (Sharkey et al. 2005).

In general, both conventional and spectral karyotyping (SKY) is unreliable in detecting copy number changes of less than 5Mb (Bar-Shira et al. 2006). Array CGH is a high resolution, whole genome based technology that has a better diagnostic detection rate of these submicroscopic chromosomal aberra­tions underlying patients with multiple anomaly syndromes; a resolution of ≤1Mb and up to 35kb has been reported (Manning & Hudgins 2007 & Bar-Shira et al. 2006). With aCGH, metaphase spreads are replaced by cloned (bacterial artificial chromosomes, BACs) or synthesized (oligonucleotides) DNA fragments across the genome that are immobilized on a glass surface, the exact chromosomal locus of which is known (Manning & Hudgins 2007). Both sample and control DNA are then hybridized onto the DNA fragments spotted onto the array; copy number variations are then measured via com­puter analysis based on the diferrences in hybridization pattern intensities.

With aCGH, the detection rate of un­balanced chromosomal rearrangements has been shown to be between 10% to 24% (Bar-Shira et al. 2006). At present, there are two types of array platform: constitutional or targeted array, and whole genome array (Manning & Hudgins 2007 & Baldwin et al. 2008). The tar­geted approach incorporates common microdeletion or duplication syndromes, telomeric regions and centromeres and selected Mendelian disorders (Baldwin et al. 2008). The development of such a custom made array has been likened to performing a ‘6000 band karyotype’. Whereas, a whole genome array has a wider coverage over the human genome. When first introduced, the whole genome array contained ~2400 BAC clones dis­tributed over the genome; the newer generation of arrays are now even more dense. The resolution of aCGH is limited by the size and distance between the DNA fragments that are spotted onto the array (Manning & Hudgins 2007). In a series by Baldwin et al. 2008, the genome-wide array coverage identified an additional 4.7% of clinically significant abnormalities that would not be detected by current targeted array platform. The average size of imbalances was ~3.7Mb and con­tained an average of 17 known genes (Baldwin et al. 2008).  In this study, ar­ray-based CGH analysis was performed using commercially available oligonuc­leotide microarrays containing about 244, 000 60-mer probes with an average res­olution of about 8.9 kb (Human Genome CGH Microarray 244A kit, Agilent Tech­nologies). The deletion was subsequently revealed by the use of array-based CGH, a molecular cytogenetic technique with an extremely high resolution. All these data confirm the importance of perform­ing a more sophisticated investigation in patients in whom a complex phenotype is strongly suggestive of the presence of a chromosomal aberration; indeed, aCGH is a powerful technique for identifying chromosomal rearrangements where conventional cytogenetics has failed.

Reports of aCGH in identifying cytoge­netic abnormalities missed by routine cytogenetic analyses have been reported in the literature and further support the sensitivity of aCGH (Shafer et al. 2006).In our patient, initial karyotyping was re­ported as normal; interpretation was however made based on a limited num­ber of metaphase spreads, as correctly cited by the laboratory. The issue of ob­taining an adequate number of meta­phase spreads occurs every now and then; this may be the result of insufficient volume of blood extracted which is not uncommon amongst paediatric patients particularly. Application of other molecu­lar technique such as Multiplex Ligation Probe-dependant Amplification (MLPA) method can improve resolution for detec­tion of changes in DNA copy number but it is typically applied in a targeted manner that assesses one or several candidate loci at a time which limit the detection rate in syndromic cases. Thus, aCGH was subsequently requested following the ‘normal’ karyotype as the patient was seen to be very ‘chromosomal’, a state­ment often concluded by clinical gene­ticists. The latter investigation indeed eluded the underlying chromosomal anomaly which was subsequently con­firmed by FISH analysis. FISH analysis is the recommended method for validating aCGH studies; furthermore, FISH is not only cost-effective but also has the ad­vantage of demonstrating the mechanism of the imbalance (Baldwin et al. 2008).

Cri-du-chat syndrome is a relatively rare syndrome and affects between 1 in 40 000 to 50 000 live births (Cornish et al. 1999). The basic defect is due to a partial deletion, either terminal or intersti­tial, of 5p15.2-p15.3. Although the name of the syndrome is derived from the high-pitched, shrill cry heard in a larger series of patients, the cry is neither pathogno­monic nor present in all patients. This anomaly is largely de novo (~85%) but may be the result of an unbalanced translocation inherited from either parent (Gorlin et al. 2001). In our patient, a 30.6Mb deletion was detected at 5p15.33-p13.3; of the 54 genes that spanned this region, 10 have been re­ported in association with clinical disord­ers. However, two of the latter have been identified to be the most significant; this was the CTNDD2 gene and SDHA gene which are associated with Cri-du-chat syndrome and Leigh syndrome. Pheno­typically, our patient showed features consistent with the former syndrome. Pa­rental karyotyping and aCGH were nor­mal, thus the deletion was de novo. Pre­natal diagnosis with FISH using uncul­tured amniocytes have been successfully reported in cases where the chromo­somal anomaly is the result of a parent being a balanced translocation carrier (Gorlin et al. 2001 & McKusick-Nathans Institute of Genetic Medicine 2010).

As anticipated, the larger the deletion, the more severe the phenotype. In addi­tion to both somatic and mental retarda­tion, patients with 5p deletion exhibit the following clinical features:  microcephaly, increased inner canthal distance (75%), down-slanting palpebral fissures (60%), epicanthic folds, broad nasal bridge with prominent nasal root and a round face. Marked hypotonia and feeding difficulties are common as seen in our patient. Other major malformations include: con­genital heart defects (30-50%), muscu­loskeletal and central nervous system anomalies [eg hypoplastic cerebellum, corpus callosum dysgenesis and arach­noid cyst] (Gorlin et al. 2001 & McKusick-Nathans Institute of Genetic Medicine 2010). The MRI brain of our pa­tient showed no specific structural ano­maly; she is presently on a multidis­cipli­nary follow-up which involves regular monitoring by the developmental Paedia­trician and Clinical Geneticist, rehabilita­tion by occupational therapy, physiothe­rapy and speech therapy, in addition to serial evaluation by audiology and oph­thalmology.

Although aCGH is increasingly used as a diagnostic tool in patients with global developmental delay, one of the main drawbacks is its potential for identifying novel copy number variations that may not be responsible for the patient’s global delay. Even if a variant is present in the affected individual but absent from “nor­mal” parental genomes, it does not nec­essarily follow that it is a pathogenetic change, and it may instead represent an innocuous copy-number polymorphism (a normal variation in the human genome). Interpretation of aCGH is often a chal­lenging process; to facilitate this process, several international databases have been established; these include DECI­PHER (Database of Chromosomal Im­balance and Phenotype in Humans Us­ing Ensemble Resources, http://www.sanger.ac.uk/PostGenomics/decipher/) and ECARUCA, the European Cytogeneticists Association Register of Unbalanced Chromosome Aberrations (http://www.ecaruca.net/). In a meta-anal­ysis by Subramonia-Iyer et al. (2007) involving seven studies amongst patients with learning disabilities, an acceptable false-positive rate (for non-causal ab­normalities) of between 5 to 10% was reported (Subramonia-Iyer et al. 2007).

CONCLUSION

Array CGH is a sensitive tool in detecting submicroscopic chromosomal aberra­tions that are undetectable by current cytogenetic tests. However, before being accepted as a routine diagnostic evalua­tion, certain issues pertaining to both clinical and laboratory aspects should be considered. These include the selection of patient phenotype, optimal array res­olution and appropriateness of array platform, establishment of quality assur­ance in a clinical set-up, availability of genetic counseling and of course, cost-effectiveness as compared to existing genetic tests.