Nicotinamide (vitamin B3) treatment improves response to G‐CSF in severe congenital neutropenia patients

Severe congenital neutropenia (CN) is an inherited bone marrow failure syndrome with markedly reduced granulocyte number and function that often manifests with life-threatening bacterial infections.1 Granulocyte colony-stimulating factor (G-CSF) treatment is the primary therapeutic approach and is aimed at keeping granulocyte levels at or above 1 000/µl to prevent infectious complications.2 Some CN patients require high doses of G-CSF to reach this goal, and a small group of patients is not responsive to G-CSF therapy at all. Overall the cumulative 15-year incidence of death from sepsis for CN patients was described to be as high as 10%.3 There is also some evidence that the cumulative risk of development of acute myeloid leukaemia (AML) or myelodysplastic syndrome (MDS) is higher in patients who require higher doses of G-CSF.3, 4 The association of G-CSF dose with the relative hazard of MDS/AML or sepsis death was also detected.3 Therefore, other therapeutic options that may lead to a reduction of G-CSF dose and increase in granulocyte functions are required. We recently demonstrated that nicotinamide (NA, vitamin B3) acts through the NAD+/SIRT1 protein deacetylation pathway, leading to an upregulating of the expression of G-CSF and G-CSF receptors.5 When we treated healthy volunteers with 20 mg/kg/day of nicotinamide (NA, Jenapharm, Jena, Germany) daily, they showed a marked increase in neutrophilic granulocyte numbers.5 Therefore, we aimed to evaluate whether the NA treatment of CN patients will increase neutrophil numbers and may lead to a reduction of the required therapeutic G-CSF dose. For this, CN patients were orally treated with 20 mg/kg/day of NA in addition to G-CSF therapy. After three months of treatment, absolute numbers of peripheral blood neutrophils, monocytes, eosinophils, thrombocytes, and lymphocytes, as well as haemoglobin, were evaluated. In some patients, blood and bone marrow counts were evaluated after one year on NA therapy. Eighteen patients were included in the study: 15 CN patients and three patients with cyclic neutropenia (CyN) (Table I). The study included 11 males and seven females at ages of 0·5 to 31 years [median 5·5 years; first quartile (Q1) 3·6, third quartile (Q3) 8·5]. The diagnosis of CN or CyN was made based on Severe Chronic Neutropenia International Registry (SCNIR) diagnostic criteria.1 Fourteen CN patients harboured mutations of ELANE, two patients had X-linked CN harbouring WAS mutations, and in two patients, no genetic defect has been detected yet. Two CyN patients harboured ELANE mutations, and in one CyN patient no mutations were detected despite clear cycling of the neutrophil counts (Table I). All patients received G-CSF in doses between 0·6 and 50·8 µg/kg/day before initiation of NA therapy. The local ethic committees approved the study protocol, the patients and\or their parents gave informed consent to participate in the study. Analysis of blood counts after three months of NA treatment revealed a gradual increase of absolute neutrophil counts (ANC) in 14 of 18 patients (78%). In one of them G-CSF was completely replaced by NA treatment (Table I). In patients on combined G-CSF and NA therapy (including mono-NA therapy), median ANC before NA was 0·95 × 109/l (Q1 0·64, Q3 1·51) and in combination with NA was 1·80 × 109/l (Q1 1·50, Q3 2·73) (Fig 1A, B), and the median increase in ANC was 0·76 × 109/l (Q1 0·64, Q3 1·4l; P < 0·001). No statistically significant differences in other blood cell counts were detected in the subsample of 16 patients for whom complete blood counts were available (Fig 1C). In nine out of 18 patients (50%), the addition of NA led to a reduction of the G-CSF dose required to reach ANCs above 1 000/µl including complete replacement of G-CSF by NA in one patient. Thus, median G-CSF dose in 18 patients was reduced from 13·3 µg/kg/day (Q1 4·4, Q3 23·8) before NA to 8·6 µg/kg/day (Q1 1·6, Q3 15·1) after three months of treatment, with a median decrease of G-CSF dose of 0·8 µg/kg/day (Q1 0, Q3 3·3; P = 0·007; Fig 1D). The beneficial effect of NA and G-CSF therapy was observed in most patients during one year of continuous treatment; one patient was lost to follow-up. After one year of observations, ANC was 1·66 × 109/l (Q1 1·00, Q3 2·85), and the median increase of ANC since the start of therapy was 0·85 × 109/l (Q1 0·08, Q3 1·57; P = 0·004). NA was well-tolerated as a long-term treatment option, without severe adverse events. Four patients have currently been treated with NA for more than five years and six patients for more than three years without exhaustion of responses to NA. During the time on combinatorial treatment with NA and G-CSF, three patients received bone marrow transplantation due to the acquisition of monosomy 7 (pt. # 5 two years, pt. #7 1·5 years and pt. # 12 five years after initiation of NA treatment). Two of these patients harboured ELANE mutations (Cys151Ser in pt. #7 and Gly214Arg on pt. #12) that are associated with high risk of AML or MDS.6 Morphological analysis in the remaining patients revealed no obvious differences in the bone marrow morphology (data not shown). We also observed a reduction of the frequency and severity of bacterial infections. As an example, one CN patient (#7) required G-CSF doses up to 50 µg/kg/day to achieve neutrophil counts above 1 000/µl. Upon addition of NA, neutrophil counts of more than 3 000/µl were achieved without further increases of the G-CSF dose. Upon continuous treatment (three months) of this patient, we were able to decrease the G-CSF dose to 12·5 µg/kg/day, with good clinical and laboratory results (Figure S1A). In one CyN patient (#18), monotherapy with nicotinamide was even sufficient to prevent infections, despite continuous cycling. This patient was treated with 5 mg/kg/day of NA for more than two years. Another CyN patient (#2) had episodes of profound neutropenia with absent neutrophils and complete unresponsiveness to G-CSF, during which he developed multiple severe infections. Upon the addition of NA, the nadir of his granulocyte level increased to above 0·40 × 109/l, with a reduction in the number of infections. Taken together, in our group of CN and CyN patients, NA treatment led to an increase of neutrophil counts and a decrease of required G-CSF doses with continuous clinical and laboratory responses. Therefore, the use of NA in combination with a reduced dose of G-CSF for the treatment of CN and CyN patients is promising and should be further investigated in a larger cohort of patients. It would be interesting to investigate whether patients with other types of inherited or acquired neutropenia will respond to NA. Particularly, implementation of NA in the treatment of chemotherapy-induced neutropenia may shorten neutropenic phases, improving chemotherapy regimens. It would also be important to investigate whether decreasing the dose of G-CSF will lead to reduced acquisition of G-CSFR mutations, a risk factor for the development of leukaemias. In three CN patients on combinatorial NA and G-CSF therapy, we observed acquisition of monosomy 7. These data suggest that reduction of the G-CSF dose may not prevent leukaemogenic transformation. However, whether NA protects from the acquisition of leukaemia-associated CSF3R and/or RUNX1 mutations or prevents leukaemogenesis has to be further investigated in a long-term observation study of a larger group of patients. Also, analysis of the granulocyte functions of NA-treated patients may further strengthen our observation of reduced frequency and severity of bacterial infections in these patients upon NA therapy. This work was supported by the BMBF (JS, KW). The authors have no relevant conflicts of interest to disclose. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.