Clinical Sequencing of 347 Children with Acquired and Inherited Bone Marrow Failure Syndromes

Background

Inherited bone marrow failure syndromes (IBMFSs) belong to a category of heterogeneous diseases in which at least one hematopoietic cell lineage is decreased in the bone marrow. They consist of more than 25 defined disease entities, including Fanconi anemia (FA), Diamond-Blackfan anemia (DBA), and dyskeratosis congenita (DC). As recent advances in clinical molecular studies have unraveled considerable pathognomonic molecular lesions in IBMFSs, the role of genetic tests has become more important in the diagnosis of these diseases. Moreover, differentiating between acquired BMFSs, including aplastic anemia (AA), and IBMFSs is very important to avoid ineffective immunosuppressive therapy and/or lethal complications of allogeneic hematopoietic stem cell transplantation.

Patients and Methods

To clarify the diagnostic efficacy of targeted sequencing with a custom capture bait, we performed targeted sequencing analysis in 347 patients with acquired BMFSs [AA, n = 164; myelodysplastic syndrome (MDS), n = 17] and IBMFSs (n = 166). The median (range) age at analysis was 7 (1-27) years. Clinical diagnoses of 166 IBMFSs were as follows: FA (n = 35), DBA (n = 31), DC (n = 13), congenital dyserythropoietic anemia (n = 9), severe congenital neutropenia (n = 7), Shwachman-Diamond syndrome (n = 6), and others (n = 65). Target capture was performed using a custom bait (SureSelect®, Agilent Technologies), covering genomic lesions for 184 genes known to be mutated in IBMFSs and associated hematological disorders, followed by massive parallel sequencing using HiSeq 2500 (Illumina).

Results

We genetically diagnosed 86 of 347 (25%) patients, including 14 with gene deletions [ FANCA (n = 5), RPS17 (n = 3), RUNX1 (n = 2), TERT (n =1), BCOR (n = 1), RPL35A (n = 1), and SPTB (n = 1)], which were detected using copy number analysis pipeline. We genetically diagnosed 60 of 166 patients (36%) with IBMFSs, including 10 whose clinical diagnoses were mismatched with genetic diagnoses provided by targeted sequencing.

Notably, we could identify diagnostic germline mutations in 10 of 164 (6%) patients with clinical diagnosis of AA; their diagnoses were revised to various IBMFSs [ FANCA (n = 2), FANCG (n = 1), TERT (n = 2), TINF2 (n = 1), RTEL1 (n = 1), RPL5 (n = 1), SBDS (n = 1), and KANSL1 (n = 1)]. Another 10 of 164 (6%) patients with AA harbored various somatic gene mutations [ RUNX1 (n = 3), TET2 (n = 2), PIGA (n = 2), BCOR (n = 1), NRAS (n = 2), KRAS (n = 1), PTPN11 (n = 1), IKZF1 (n = 1), SETBP1 (n = 1), IDH1 (n = 1), WT1 (n = 1), ZRSR2 (n = 1), and EPB42 (n = 1)]. Taken together, our targeted sequencing analyses revealed germline/somatic mutations in 20 of 164 (12%) patients with AA and somatic mutations in 6 of 17 (35%) patients with MDS.

Discussion

Our targeted sequencing analysis achieved a satisfactory diagnostic rate and supported the effectiveness of massive parallel sequencing as a diagnostic tool for patients not only with IBMFSs but also acquired BMFSs.