Clinical and Molecular Heterogeneity of Moderate Aplastic Anemia

Based on clinical Camitta criteria, acquired aplastic anemia is categorized according to the degree of blood count depression as severe (sAA) or moderate (mAA). In some cases mAA is a precursor of sAA; in others it is a pathophysiologically distinct chronic non-progressive entity (chronic mAA (cmAA)). It is also possible that some seemingly mAA cases represent entities ranging from congenital bone marrow failure syndromes or hypocellular MDS to misdiagnosed systemic conditions with secondary aplasias. Discriminating true cmAA from these may be important for clinical management and requires prolonged observation.

We have been able to accurately dx and treat patients (pts) with acute sAA. In contrast, pts with cmAA can have mildly depressed counts, which remain relatively stable for yrs without the need for treatment (tx).Tx delay, though, may lead to unopposed destruction of stem cells.

At our institution we have evaluated and managed 308 AA pts from 1998-2018. Of these pts, we identified 88 who met the Camitta criteria for mAA and of these, 2 progressed to sAA within 3 mos, 1 had clinically significant PNH at dx, and 85 were truly cmAA. Focusing on this true cmAA cohort our goals were to identify its distinct clinical features and response to therapies.

The median f/u for the cohort was 45 mos, with median Hgb 10.1 g/dl, ANC 1.36 k/uL, Plts 47 k/uL, and absolute retic count of 0.053 M/uL at dx. The median age for cmAA was 43 yrs (range 6-88), with 55% (47/85) females (vs 48% (108/223) for sAA p=.31). By ultra-sensitive flow cytometry, PNH clones were found in 26% (22/85) of pts at presentation of cmAA vs 22% (50/223) for sAA. From initial dx to last f/u 72% (61/85) remained mAA, of whom 57% (35/61) of pts did not require any tx. Those pts who required tx received at least one line of therapy, in most instances immunosuppression with cyclosporine (ORR~67%) and supportive care. Of those who required tx, 46% achieved a CR, 35% PR and 19% NR to 1^st line tx.

As no obvious distinction for cmAA was identified based on clinical presentation, we next classified pts based on the number of cell lineages involved, transfusion requirements, and severity of the counts (characterized into mild, moderate and severe based on severity of deviation from normal). At the time of dx 29 pts were transfusion dependent, of which one pt had a single cytopenia, 8 pts had bicytopenia, and 20 had pancytopenia. Based on our findings 34% (10/29) of pts who were transfusion dependent with borderline severe pancytopenia received ≥ 1 therapy and eventually progressed to sAA, in contrast to those mAA that were transfusion independent with similar counts. The overall median time to transformation was 14 yrs (CI=95%, 5.9-21.7). Average time to progression to sAA was 31 mos, with a progression rate of 21% (CI=95%; 10-30). Across the entire cohort, 16% (14/85) progressed to sAA, 11% (9/85) to full-blown PNH, and 1% (1/85) to AML. At progression 3 pts acquired -7 and another acquired additional chromosome six. One pt did receive allo-HSCT for mAA. In contrast, in sAA 8.5% (19/223) pts progressed to MDS/AML and 8.5% (19/223) to PNH, where the median time to transformation was 24 yrs.

We also performed a sub-cohort analysis of somatic events present in these pts: at dx 79% (67/85) had normal karyotype. NGS for somatic mutations revealed the presence of at least 1 mutation in cmAA and sAA (p:0.47). Serial NGS was available for 11/85 for assessment of clonal dynamics. At the time of dx 8 out of 11 pts had no mutations at presentations and the three other pts had a single mutation in RUNX1, ASXL1, and PIGA, of which the RUNX1 was found to be a transient clone (VAF: 30%) at the time of last f/u. The pt with ASXL1 eventually progressed to MDS with -7 with expansion of the clone from 3 to 19% and the pt with a PIGA clone was never treated for mAA progressed to full-blown PNH requiring tx with eculizumab. NGS from the last follow up revealed 6/11 pts acquired single mutation ETV6 (VUS), ASXL1, and PIGA, of which 4 pts went on to receiving tx for PNH.

In sum, our results suggest that cmAA does not have, unlike sAA, a relentless clinical course. The rate of evolution from mAA to sAA was 30% in 10 yrs with median time to progression 2.6 yrs, with a similar rate of evolution to PNH with median time to progression of 4yrs and only one pt progressed to AML. In relation to sAA, there was not a significant difference in progression to MDS or PNH and the survival and response to therapy of these pts was excellent as expected from Camitta's criteria.