Pancytopenia with a severe decrease in bone marrow (BM) cellularity in children may be caused by a broad variety of underlying disorders. The appropriate classification of bone marrow failure syndromes in children is challenging, particularly with respect to the histological distinction between aplastic anemia (AA), refractory cytopenia of childhood (RCC), and refractory cytopenia with multilineage dysplasia (RCMD). Inherited bone marrow failure syndromes (IBMFS) further defy accurate diagnosis. Clonal hematopoiesis in AA is indicated by the presence of paroxysmal nocturnal hemoglobinuria (PNH) cells and by the identification of uniparental disomies involving chromosome 6p. In addition, "clonal transformation," as defined by the development of myelodysplastic syndromes (MDS) or acute myelogenous leukemia has been noted in about 15% of AA patients. In adult patients with AA, somatic mutations were frequently detected in myeloid malignancy-related genes such as DNMT3A, BCOR, and ASXL1. We aimed to characterize the genetic background of childhood AA/RCC/RCMD.

Patients and Methods

We studied 168 patients with idiopathic AA/RCC/RCMD in children. Diagnosis with AA, RCC, and RCMD was made on the basis of the 2008 World Health Organization classification criteria. Blood, bone marrow, and buccal samples were obtained from the patients after written informed consent was received according to protocols approved by the ethics committee of Nagoya University Graduate School of Medicine. Target sequencing (n = 168) was performed for 88 IBMFS-associated genes and 96 myeloid malignancy-related genes. Furthermore, whole-exome sequencing (WES, n = 25) was performed with matched tumor/normal samples. The mean depth for targeted sequencing was 451x, and the mean depth for WES was 103x. Somatic mutations were detected with the use of a frequency threshold of 0.07 (WES) or 0.02 (targeted sequencing) for variant allele frequency and were individually validated with the use of deep sequencing of polymerase-chain-reaction-amplified targets.


Only one germline mutation that was diagnostic of IBMFS was detected in our cohort (0.6%). It was a RTEL1 mutation, which supported the diagnosis of dyskeratosis congenita. Telomere length of the patient with a RTEL1 mutation was shorter compared with that of age-matched healthy individuals (−3.2 Standard Deviation). WES, performed in 25 patients, detected only three somatic mutations, all of which affected BCOR. In target sequencing, 20 somatic mutations were detected in 19 patients (11.3%). BCOR (n = 9) and PIGA (n = 4) were recurrently mutated. The mutational frequency of DNMT3A and ASXL1 was very low (0.6%) in our cohort and was clearly different from that of an adult cohort. The majority of somatic mutations carried low variant allele frequency. In case of BCOR mutations, the variant allele frequency tended to be low, suggesting subclonal composition. In case of U2AF1 mutations, the variant allele frequency tended to be high, which suggests that the U2AF1 -mutated clone was dominant in the bone marrow. The difference in the frequency of somatic mutations in AA, RCC, and RCMD was not statistically significant (p = 0.49). However, with regard to the mutated genes, two patients with RCMD carried U2AF1 plus SETBP1 and TP53 mutations, respectively, which are well-known predictors of poor prognosis in adult MDS. The patient who carried U2AF1 plus SETBP1 developed MDS later and underwent bone marrow transplantation. Of the 19 patients with mutated genes, 15 patients were treated with immunosuppressive therapy (IST). The response rate to IST at 6 months was 60% in the patients with somatic mutations, which was equivalent to those without mutations.


In our cohort of children who were clinically diagnosed with AA/RCC/RCMD, the frequency of cryptic IBMFS was considered low. Furthermore, the frequency of detectable somatic mutations in childhood AA was low compared with that in adult AA. No novel mutational target was identified with WES. Idiopathic bone marrow failure syndromes in children were characterized by a paucity of gene mutations irrespective of the histopathological classification. Mutations in adult MDS-related genes suggest molecular pathogenesis is different between RCMD and AA/RCC. In conclusion, our study clarified the yet unrevealed genetic background of idiopathic bone marrow failure syndromes in children.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.