Response to Letter to the Editor from Youn Hee Jee: "Familial Short Stature - A Novel Phenotype of Growth Plate Collagenopathies".

In the Letter to the Editor (1) commenting on our article entitled “Familial Short Stature - A Novel Phenotype of Growth Plate Collagenopathies” (2) by Youn Hee Jee, the results of the interpretation of genetic variants have been questioned based on the different variant classification in the ClinVar database.

ClinVar is a freely available, public archive of human genetic variants and interpretations of their significance to disease (3). Importantly, the database also includes misclassified variants that are frequently reclassified if evaluated manually, especially in cases where the gold star (review) status in ClinVar is low (4, 5). The result of the evaluation of variants in ClinVar without further knowledge should, therefore, not exclude the identified variant from further considerations of pathogenicity; in contrast, each variant should be reevaluated when new facts are discovered or presented. For that reason, we and other authors use ClinVar as only one source of information about the identified variants; the final classification is always made using American College of Medical Genetics and Genomics standards and guidelines as recommended (6). The most important criteria for variant classification includes multiple in silico programs predicting the deleteriousness of the variant, location of the variant in a functional domain, the frequency of the variant in population databases and, uniquely in our study, the segregation of the variant with short stature and other clinical signs associated with collagenopathies within the family, which was frequently the most important criterion for correct variant classification. The prediction of pathogenicity of variants should therefore be evaluated by an assembly of multiple tests with clinical data, in accordance with the principles of translational research.

The example of the conflicting variant interpretation in our study and in the ClinVar database noted by Dr Jee is the variant c.1300C > T (p.Pro434Ser) in the COL2A1 gene. This variant causes a substitution of proline with serine in the triple-helical region of the gene that affects the thermal stability of the helix (7); it is predicted to be pathogenic by in silico models (Mutation Taster, PolyPhen, SIFT, CADD), and it segregated well with extremely short stature in 3 generations of the family (for details see the supplementary materials of the original article). Moreover, the variant was previously described in the literature in a patient with Kniest dysplasia, a condition typical of causative variants in the COL2A1 gene (7). All of the aforementioned factors indicate the causality of the variant. The ClinVar classification of the variant as benign was based solely on a single article evaluating genetic variants relating only to hearing loss but not to short stature or clinical signs of bone dysplasia (8); therefore, it is ranked with only the lowest score of one star in the ClinVar reliability classification. Similarly, we believe that all other variants mentioned in the Letter to the Editor have been classified correctly, and the proof of their causality is well documented in the original article and its supplementary materials (2).