New classification of Cutis Laxa types

Dr A.Beyens presented the long piece of work that analized the different symptoms of all Cutis Laxa types. The files from the 192 patients of the in house cohort  studied by Pr Bert Callewaert’s team as well as the cases published in literature were screened by Artificial Intelligency (AI) in order to group major symptoms from all types.

623 files were thus analysed. Clinical and molecular features, plus 290 parameters were taken into account which allowed AI to define 6 groups which share clinical ad pathophysiologic features : Skin, Cardiovascular, Pulmonary, Bone, Glycosylation and Mithocondrial. Moreover, these groups show significant differencies in the elastic fibers degradation.

Thanks to this study, it is now possible to build a decision tree (CART Algorithm) that, depending on the clinical and pathophysiological signs, can determine a suspicion of the type of Cutis Laxa from which the patient is suffering.

The results obtained have a reliability of nearly 90%. This does not diminish the need for genetic/molecular testing to confirm this suspicion. But, it allows to immediately organize the care and follow up needed depending on the type, without waiting for the 6 to 8 months needed for the genetic test results. It is a significant improvement to diagnostic delay and improved care for the patients.