Hepatitis-associated aplastic anemia (HAAA) is a rare variant of bone marrow (BM) failure that is typically diagnosed 2-3 months after an acute attack of hepatitis in patients that are seronegative for known hepatitis viruses. Although very little is known about its etiology, an autoimmune mechanism is presumed based on clinical response to immunosuppressive therapy (IST) and reports of oligoclonal T-lymphocyte expansion and skewing of CDR3 region length. However due to the rareness of this disease, the published evidence is still limited.

We analyzed 18 pediatric patients consecutively diagnosed with HAAA in the Czech Republic between 2004-2019. Based on BM biopsy examination, 14 patients had aplastic anemia (AA) and interestingly 4 patients were histologically consistent with refractory cytopenia of childhood (RCC). Median age of these patients was 6.6 years (range 1-18) with a male to female ratio of 2:1. The prevalence of HAAA in our cohort of pediatric patients with BM failure (n=107; excluding patients with known genetic cause) was 17%. The presence of a paroxysmal nocturnal hemoglobinuria clone was excluded and cytogenetic evaluation showed no abnormalities.

We examined both the diagnostic BM and peripheral blood by flow cytometry in all patients (with the exception of one peripheral blood sample) and performed deep immunophenotyping of T cells in 7 of them. Ten patients were selected for T cell receptor beta (TRB) repertoire and whole-exome sequencing (WES) based on the availability of DNA samples isolated from the BM before the start of the treatment. TRB sequencing was performed following the SOP developed by the EuroClonality-NGS working group with DNA input normalized to the equivalent of 20 000 CD3+ cells per sample based on flow cytometry data. Bioinformatic analysis was performed with ARResT/Interrogate. WES was performed on Illumina NextSeq 500 and libraries prepared using the Agilent SureSelectXT Human All Exon V6+UTRs kit. Vertebrate virus detection was performed using the VirCapSeq-VERT probe capture enrichment set and massive parallel sequencing from 8 BM DNA samples taken at the time of diagnosis.

We found significant activation of CD8 positive T cells based on the expression of HLA-DR compared to non-HAAA RCC and AA patients (n=20 and n=21 respectively, Mann-Whitney p<0.0001). Three out of 7 patients with deep T cell immunophenotype available showed a decrease of naive CD27pos45RAposCD8pos T cells. From TRB repertoire analysis we were able to identify a total of 5 private immunodominant clones in 3 patients with relative abundance of up to 11% from all productive TRB sequences. These clones were not found in any previous publications or clonotype databases based on their CDR3 sequence. We detected 5 clones shared by 3 or more patients; these were previously published in other cohorts and are not specific for HAAA. WES analysis did not reveal any previously published variant in genes associated with immune dysregulation or immunodeficiency, or any candidate variant for functional validation. No relevant vertebrate virus signal was detected in the 8 samples.

Out of all 18 patients, 3 patients underwent MSD-HSCT without preceding immunotherapy. Fifteen patients were treated according to standard European Working Group on MDS/SAA protocols with anti-thymocyte globulin (horse n=6; rabbit n=9), corticosteroids and cyclosporin A. In total, 7 patients reached complete remission on IST by day 120 and 7 patients with insufficient response underwent MUD-HSCT. One patient died shortly after MUD-HSCT because of toxicity. We did not observe any correlation of T cell populations at diagnosis or day 120 of IST (activated and exhausted T cells) and clinical response. Similarly, the presence of an immunodominant clone or oligoclonal TRB repertoire with lower diversity at diagnosis had no significant impact on the outcome of these patients in our cohort.

In conclusion, we present the biggest cohort of pediatric patients with HAAA analyzed by next-generation sequencing and flow cytometry so far. Even though our results show significantly increased activation of CD8 positive T-cells and presence of expanded clones, we did not confirm that these factors would be useful and significant for the prediction of treatment outcome. Large number of our patients did not fully respond to IST, requiring HSCT.

Supported by AZV 18-07-00430, 16-32568A and PRIMUS/17/MED/11.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution