A diagnostic protocol for autism: an Italian experience

Background and aims: The usefulness of a protocol for aetiologic diagnosis in pervasive developmental disorders (PDD) has been considered in literature and the results are heterogeneous. We addressed this topic, evaluating an Italian sample of patients with PDD. Material and methods: We included in the study the patients referred to the Autism Centre of the University of Bologna from January 1999 to September 2009, affected by PDD according to DSM-IV-TR. All the subjects underwent anamnesis, clinical, laboratory and instrumental investigations for an aetiological diagnosis. First, we evaluated neurobiological findings in the whole sample. After, we divided the patients into two groups, respectively syndromal (with genetic or metabolic syndromes, cerebral lesions and/or microcrania/macrocrania) and nonsyndromal (without genetic, metabolic diseases, cerebral lesions and/or microcrania/macrocrania), that we compared each other for several variables. Results: We selected 90 subjects (mean age 11 years 1 month, range 2-36 years, mean follow-up 3 years 8 months), respectively with autistic disorder (55.5%), pervasive developmental disorder not otherwise specified (38.9%), Asperger disorder (4.4%), Rett disorder (1.1%). Male-to-female ratio was 3:1. Organic pre-, peri-, and postnatal antecedents were found in 27.8% of subjects. Neurological signs were present in 83.3%; brain imaging showed pathological findings in 33.3%. Genetic examinations disclosed abnormal karyotype in 7.8%, one patient had the Cowden syndrome, one the Rett disorder variant CDKL5, one the Wilson-Turner syndrome. The syndromal group consisted of 42 subjects (46.7%), the non-syndromal one of 48 patients (53.3%). Dysmorphisms prevailed significantly in the syndromal cases (30.9%), but they were present also in 10.4% of the non-syndromal ones. Severe/profound mental retardation recurred more often in syndromal cases respect to non-syndromal ones (42.9% vs 29.2%), but the difference was not significant. We found epilepsy in 30.0% of the whole sample without a significant group difference. Conclusion: The identification of specific pathologies underlying PDD could help to provide a medical therapy in tractable conditions, and to give a genetic counselling. In our sample genetic examinations, MRI and EEG recordings had the best diagnostic yield, while metabolic tests and other examinations were negative or mildly altered. Diagnostic evaluation should be performed choosing the examinations based on clinical suspicion with minimal resources and patient discomfort and with the best diagnostic yield.