Array analysis and karyotyping: workflow consequences based on a retrospective study of 36,325 patients with idiopathic developmental delay in the Netherlands.

Item does not contain fulltext
Anomalies of chromosome number and structure are considered to be the most frequent cause of unexplained, non-syndromic developmental delay and mental retardation (DD/MR). High-resolution, genome-wide, array-based segmental aneusomy profiling has emerged as a highly sensitive technique for detecting pathogenic genomic imbalances. A review of 29 array-based studies of DD/MR patients showed that a yield of at least approximately 19% pathogenic aberrations is attainable in unselected, consecutive DD/MR referrals if array platforms with 30-70 kb median probe spacing are used as an initial genetic testing method. This corresponds to roughly twice the rate of classical cytogenetics. This raises the question whether chromosome banding studies, combined with targeted approaches, such as fluorescence in situ hybridisation for the detection of microdeletions, still hold substantial relevance for the clinical investigation of these patients. To address this question, we reviewed the outcome of cytogenetic studies in all 36,325 DD/MR referrals in the Netherlands during the period 1996-2005, a period before the advent of array-based genome investigation. We estimate that in a minimum of 0.78% of all referrals a balanced chromosomal rearrangement would have remained undetected by array-based investigation. These include familial rearrangements (0.48% of all referrals), de novo reciprocal translocations and inversions (0.23% of all referrals), de novo Robertsonian translocations (0.04% of all referrals), and 69,XXX triploidy (0.03% of all referrals). We conclude that karyotyping, following an initial array-based investigation, would give only a limited increase in the number of pathogenic abnormalities, i.e. 0.23% of all referrals with a de novo, apparently balanced, reciprocal translocation or inversion (assuming that all of these are pathogenic), and 0.03% of all referrals with 69,XXX triploidy. We propose that, because of its high diagnostic yield, high-resolution array-based g