Genotype-Phenotype Associations in Patients with Severe Congenital Neutropenia.

BACKGROUND: G-CSF therapy has reduced sepsis mortality in patients with severe congenital neutropenia (SCN), revealing a syndromic predisposition to myelodysplastic syndrome and acute myeloid leukemia (MDS/AML). The MDS/AML risk appears to be higher in SCN patients who are less-responsive to therapy; however, the genetic determinants of susceptibility remain to be elucidated.

METHODS: We studied 82 North American and Australian patients with SCN on long-term G-CSF enrolled in the Severe Chronic Neutropenia International Registry. These patients had prospective follow-up, sufficient banked DNA for genotyping, and validated clinical data.

RESULTS: Fifty-two patients (63%) were positive for germline mutations in neutrophil elastase (ELA2); the remaining 30 patients (37%) had no detectable mutations. Prior to G-CSF therapy, blood cell counts were significantly different between ELA2 positive and negative patients. Compared with ELA2 positive patients, ELA2 negative patients had significantly higher median absolute neutrophil count (ANC) values (158 versus 53 cells/uL, P=0.02), significantly lower monocyte and eosinophil counts (P<0.001), and significantly lower platelet counts (P=0.002). ELA2 negative patients were maintained on significantly lower doses of G-CSF (5.2 versus 9.9 ug/kg/day, P=0.02), consistent with higher baseline ANC values. However, ELA2 negative patients were significantly less responsive to G-CSF therapy. Compared with ELA2 positive patients, on therapy, ELA2 negative patients had significantly smaller median increases in ANC values (1200 versus 2700 cells/uL, P=0.02); the difference remained significant after controlling for individual G-CSF dose (P=0.04). ELA2 negative patients had 5 MDS/AML events in 132 person-years, versus 8 MDS/AML events in 373 person-years among ELA2 positive patients. Using the Cox proportional hazards model, the hazard of MDS/AML was 3.1-fold higher in ELA2 negative patients compared with ELA2 positive patients. This association was of borderline statistical significance (P=0.08). No patient in either group died of sepsis. We found 34 distinct mutations in 52 ELA2 positive patients. The mutation sites were not clustered in the 8 ELA2 positive patients who developed MDS/AML. We found no associations of phenotype or outcome with specific ELA2 mutations; however, the numbers were limited.

CONCLUSIONS: SCN patients without mutations in ELA2 constitute a clinically distinct subgroup. ELA2 negative patients are significantly less responsive to G-CSF therapy than ELA2 positive patients, and they may be at substantially higher risk of leukemia. Additional studies are needed to confirm the latter association and identify the causative gene or genes.