ELANE Mutations in Cyclic and Congenital Neutropenia: Genotype-Phenotype Relationships,

Abstract 3398

Mutations in ELANE are the most common cause of cyclic and congenital neutropenia. We have detected 65 distinctive and presumably causal mutations in 212 patients with cyclic (91) and congenital neutropenia (121), including 28 families with a total of 101 affected members, followed longitudinally through the Severe Chronic Neutropenia International Registry and Repository. The diagnosis of cyclic neutropenia was associated with 13 different mutations. Congenital neutropenia occurred with 61 different mutations and 9 mutations were found in both cyclic and congenital neutropenia patients.

ELANE mutations causing cyclic neutropenia cluster at the junction of exons 4 and 5 and include mutations in intron 4 (83/91). In a majority of the patients (64/91), the mutations are predicated to alter the binding pocket for the enzyme with its substrates based on application of the tertiary structure of neutrophil elastase using ICM-Browser-Pro software (Molsoft, San Diego, CA). However, the proportion of specific mutations affecting the binding pocket was similar for cyclic neutropenia 6/13 versus congenital neutropenia 29/61 (P=1, Fisher exact test). For the unique mutations associated with cyclic neutropenia, 2/4 mutations were predicted to affect the binding pocket.

ELANE mutations causing autosomal dominant severe congenital neutropenia have been detected in exons 1, 2,3,4,5, and introns 3 and 4, with a wide diversity of predicted effects on the structure of mutant neutrophil elastase. An equivalent proportion of cyclic and congenital neutropenia patients affect the two glycosylation sites (congenital 55/121 versus cyclic 49/91, p=0.26, Fisher exact test). There was also an equivalent proportion of specific mutations in the two groups (cyclic 7/13, congenital 29/61, p=0.77, Fisher exact test).

Seven of 7 mutations predicted to affect the structure of the active site of the enzyme caused congenital neutropenia in 10 patients (mutations exon 2: C26S, C42F, C42S; exon 4: Del 145–152, Del 161–170 2fs; exon 5: Del G172 fs182ter, C194ter). A significantly higher proportion of patients with active site mutations evolved to develop leukemia or myelodysplasia (AML/MDS) (4/10 active site mutations vs. other ELANE mutations 16/111, p<0.05, Fisher exact test), There was no difference in the age, age at the initiation of G-CSF therapy or the daily G-CSF dose for this cluster of 10 patients.

As previously reported (Rosenberg et al Br J Haematol. 2008; 140: 210 and 2010; 150: 196) patients with congenital neutropenia who require higher doses of G-CSF are at greater risk of AML/MDS, whether or not they have ELANE mutations. We have not identified specific mutations associated with a higher risk of leukemia. There were no cases of AML/MDS in the 91 patients with cyclic neutropenia, but there were 20 cases of AML/MDS in the population of 121 congenital neutropenia patients. Two congenital patients with G185R mutations evolved to develop AML/MDS; 5 others with this same mutation have not shown signs of evolution. There are 18 different mutations in the other 18 patients. In the congenital neutropenia patients, we have found no apparent difference in the risk of AML/MDS for mutations affecting the binding pocket or glycosylation sites of the enzyme. In this group, we also cannot associate a requirement for higher G-CSF doses or failure to respond with specific mutations or categories of mutations. The distinctive features of ELANE mutations causing cyclic neutropenia, refractoriness to G-CSF and the risk of leukemia are still unknown.