LAUSR

We appreciate the comments by Lucijanic and Kusec on our article “Mutational profile and haematological response to iron chelation in myelodysplastic syndromes (MDS)” (Fabiani et al, 2019). Following their suggestions, we further investigated the mechanisms of haematological response to iron-chelating therapy (ICT) in MDS, by screening for haemochromatosis gene (HFE) single nucleotide variants (SNV) in 44 patients for whom genomic DNA was available. HFE polymorphisms are relatively common in Caucasians. The frequency of the major HFE polymorphism, H63D variant is around 22–28%, whereas those of S65C and C282Y are lower (6–10%) (Marchall et al, 1999; De Souza et al, 2015; Lee et al, 2017). HFE SNVs H63D, S65C and C282Y were analysed by Sanger sequencing using the BigDye Terminator v.3.1 cycle sequencing kit (ABI PRISM 3100; Applied Biosystems/Life Technologies, Milan, Italy). Primer sequences were HFE (H63D and S65C) forward 50-GTTCACACTCTCTGCACTACCTC-30, HFE (H63D and S65C) reverse 50-CTTGCATTTCACAGCCCAGGATGACC-30, HFE (C282Y) forward 50-GCCTTGAACTACTACCCCCAGAACATC-30 and HFE (C282Y) reverse 50ACTCACTTGAACCCTGCCTCTTCC-30. All identified SNVs were confirmed in an independent experiment. Hardy–Weinberg equilibrium was calculated for each SNV using the Pearson v test. Differences in genotypes distribution between resistant and responsive patients were evaluated using v test (Yates corrected). Odds ratios with 95% confidence intervals were also calculated for the genotypes. For expected cell values less than 5, the Fisher’s exact test and the limits for confidence intervals were preferred. All statistical tests were performed using GraphPad Prism version 6.0 g (GraphPad Software, Inc., San Diego, CA, USA). Patient’s characteristics have been previously reported (Fabiani et al, 2019). The frequency of HFE SNVs in our patient cohort was as previously reported in the literature (Marchall et al, 1999; De Souza et al, 2015; Lee et al, 2017). In particular, the frequency of the H63D variant was 22 7% (10 out of 44 patients), whereas it was 4 5% (2 out of 44 patients) for both S65C and C282Y SNVs. Both H63D and S65C were present in one patient. In our study population, no significant differences in HFE H63D genotype frequency were found in patients who achieved haematological improvement during ICT (4 out of 15 patients, 26 6%) vs non-responders (6 of 29, 20 7%). The frequency of HFE S65C and HFE C282Y variants was too low (4 5% for each) to assess the role of these polymorphisms in the context of ICT, thus formal analysis included only the HFE H63D variants. Ferritin levels are frequently increased in patients with MDS at the time of initial diagnosis, prior to red blood cell transfusions, probably as a result of ineffective haematopoiesis and of the inflammatory environment, and have been shown to play a significant prognostic role in these patients (Voso et al, 2013). However, the increased ferritin levels at enrolment in our study were the consequence of iron overload due to transfusion therapy, the burden of which was variable in these patients. Accordingly, iron accumulation, as mirrored by ferritin concentration, was not associated with HFE genotype in our patient cohort: ferritin level at the time of treatment start was similar in carriers of the HFE H63D variant and wild-type patients (average 1699 vs. 2138 lg/l, respectively; P = 0 237). Also, there were no differences in absolute ferritin levels (mean 1371 vs. 1450 lg/l, respectively; P = 0 869), or their relative decrease after 6 months of ICT in HFE genotype carriers, versus wild-type patients. As suggested by Lucijanic and Kusec, we also looked at the correlations between the prevalence of somatic mutations in our patient cohort and HFE SNVs. We found no significant correlations between number of somatic mutations and HFE genotypes (H63D, S65C and C282Y), although the numbers might be too small to detect significant differences. Also, we did not find any significant associations between HFE variants and other patient characteristics, including morphology and revised international Prognostic Scoring System variables, such as karyotype, bone marrow blast proportion and peripheral blood cytopenias. In conclusion, with the limitations of a small patient cohort, our data show that HFE H63D SNVs do not play a role in the haematological response to iron chelation therapy in patients with lower-risk MDS. However, analysis of nonHFE H63D variants was not performed and therefore cannot be excluded as having possible associations with ferritin level, response to iron chelation and/or myeloid neoplasm-related gene mutations. Further studies in larger patient cohorts may be useful to address the role of these rare variants.