Distinctive and common features of moderate aplastic anaemia

Abstract

The therapy algorithm for severe aplastic anaemia (sAA) is established but moderate AA (mAA), which likely reflects a more diverse pathogenic mechanism, often represents a treatment/management conundrum. A cohort of AA patients (n = 325) was queried for those with non-severe disease using stringent criteria including bone marrow hypocellularity and chronic persistence of moderately depressed blood counts. As a result, we have identified and analyzed pathological and clinical features in 85 mAA patients. Progression to sAA and direct clonal evolution (paroxysmal nocturnal haemoglobinuria/acute myeloid leukaemia; PNH/AML) occurred in 16%, 11% and 1% of mAA cases respectively. Of the mAA patients who received immunosuppressive therapy, 67% responded irrespective of time of initiation of therapy while conservatively managed patients showed no spontaneous remissions. Genomic analysis of mAA identified evidence of clonal haematopoiesis with both persisting and remitting patterns at low allelic frequencies; with more pronounced mutational burden in sAA. Most of the mAA patients have autoimmune pathogenesis similar to those with sAA, but mAA contains a mix of patients with diverse aetiologies. Although progression rates differed between mAA and sAA (P = 0·003), cumulative incidences of mortalities were only marginally different (P = 0·095). Our results provide guidance for diagnosis/management of mAA, a condition for which no current standard of care is established.

Keywords: acute myeloid leukaemia; clinical outcomes; moderate aplastic anaemia; molecular mutation; myelodysplastic syndrome.

Figure 1

Clinical course, treatments and response to therapy. (A) The horizontal bar graph represents individual pts with moderate aplastic anaemia (mAA) with their respective clinical course (n = 85) and the start of mAA therapy is depicted by the black triangle. The pie chart depicts the proportions of pts with mAA and those that transformed to severe aplastic anaemia (sAA), PNH and AML. (B) Clonal dynamics of mAA versus sAA: at diagnosis, during clinical course, and last follow‐up prior to progression. (C) The pie diagrams indicate 1st (left) and 2nd (right) lines of treatment. The bar graphs depict the individual responses to therapy. Supportive care not included (transfusions, growth factors, and antibiotics). ATG, anti‐thymocyte globulin; CsA, ciclosporin; HSCT, haematopoietic stem cell transplant; NR, no response; PR, partial response; CR, complete response.

Figure 2

Cumulative incidence of progression and mortality in moderate (mAA) and severe (sAA) aplastic anaemia. (A) Cumulative incidence (CI) of progression in mAA and sAA. (B) CI of mAA to sAA. (C) CI of paroxysmal nocturnal haemoglobinuria (PNH) between mAA and sAA. (D) CI of MDS/AML in mAA and sAA. (E) CI of mortality in the mAA patients (non‐progressors n = 61) (F) CI of mortality in mAA and sAA.

Figure 3

Diagnosis and treatment algorithm. (A) Pathogenic and diagnostic overlap of mAA. (B) Treatment algorithm for AA with moderate counts: chronic moderate aplastic anaemia (cmAA) without clinical progression and progressive moderate aplastic anaemia (pmAA) with clinical presentation similar to that of sAA. BMF, bone marrow failure; Hgb, haemoglobin; ANC, absolute neutrophil count;