Aplastic Anemia in the Elderly: Treatment Patterns and Outcomes from an International Observational Study

Aplastic anemia (AA) is a rare bone marrow failure disorder characterized by a bimodal incidence with a peak in the elderly >60 years. While it has been known since a long time that the outcome of immunosuppressive therapy and hemopoietic stem cell transplant (HSCT) is inferior in the elderly, little evidence is available since the introduction of novel treatment modalities such as eltrombopag. With the aim of deciphering treatment and outcomes of elderly AA, we accrued AA patients diagnosed according to Camitta criteria at 4 tertiary hematologic centers in Italy (Milan/Rome), USA, and UK between 1976-2024. Details on baseline hematologic features, bone marrow evaluation, and mutational status by NGS of common myeloid drivers were reviewed. The different management strategies were registered and categorized as: 1) cyclosporin (CyA), 2) eltrombopag + CyA, 3) eltrombopag, 4) ATG, and 5) HSCT. Responses were assessed according to EBMT criteria (complete, CR, if platelets PLT >100 x 10^9/L, hemoglobin Hb >10 g/dL, neutrophils-ANC > 1.5x10^9/L; partial if transfusion independence). All data were analysed according to age categories <60 vs > 60 years (namely young vs elderly AA); the latter were further divided into 60-64 vs 65-69 vs >70 years categories for continuous variables analyses.

Overall, 1110 patients were included: 319 (29%) patients were >60 years old (N=85 had 60-64 years, 106 65-69 years, and 128 >70 years category) and 791 (71%) <60 years old at diagnosis. Male to female ratio (1:1) and distribution of different severity categories was comparable across ages (overall, 47% non-severe AA, 44% severe AA, and 9% from very severe AA). The severity of thrombocytopenia at presentation was greater in elderly vs younger patients (mean 57 in <60 years, 43 in 60-64, 40 in 65-69, and 30x10^9/L in >70 years, p=0.001), whilst mean Hb (9.2, 9.3, 9.4, 9.1 g/dL) and neutrophil levels (1, 1.2, 1.2, 1x10^9/L), as well as the rate of red cell and platelet transfusions (83 in <60 vs 79% in >60 years and 84 vs 76%, respectively) were comparable. A PNH clone was detected in 43% of patients with similar frequency across ages, but with a smaller size on granulocytes in elderly patients (mean 15% in <60 years, 8% in 60-64, 6.4% in 65-69, and 6.3% in >70 years, respectively, p=0.002). Furthermore, elderly patients more frequently presented cytogenetics aberrations (29/319, 9% vs 34/791, 4%, p=0.004), none MDS defining, whilst the frequency of somatic mutations by NGS myeloid panel (21%) was comparable among age subgroups, as did the prevalence of bone marrow reticulin fibrosis (WHO grade 1, 13%).

Regarding treatment patterns, 4.3% of >60 years old patients and 8% of <60 years received no treatment, while the remaining received ATG plus CyA (43% vs 59%), CyA only (30% vs 15%, p<0.0001), CyA plus eltrombopag (7% vs 3%), or eltrombopag only (4% vs 1%). Elderly patients showed a significantly lower overall response (complete + partial) to first-line therapy at 6 months as compared to younger patients (47% vs 65%, p<0.0001), while the rate of adverse events was comparable across age categories (10% overall). No elderly patient underwent frontline transplant (0 vs 9% of patients <60 years, p<0.001), whilst 27 (8%) received transplant as second line therapy (vs 19% of younger cases, p<0.0001). Overall, 12% of patients required eculizumab therapy for their PNH clone, less commonly in the elderly subgroup (6% versus 13%, p=ns).

The rate of evolution to MDS was greater in elderly patients (12% vs 6% in younger AA, p=0.002), whilst the rate of evolution to AML was similar across ages (1.2 vs 1.3%); 6% of elderly and 11% of younger AA patients (p=0.01) developed a clinically significant PNH clone requiring anti-complement treatment. During the follow up (median 12 years) 268 patients (24%) died, with an increasing frequency across age categories (15% in <60 years, 27% 60-64, 45% 65-69, and 61% >70 years). The following hazard ratios for age-related mortality emerged as compared to younger patients: 1.97 (95% CI 1.2-3) for 60-64 years, 4 (2.8-5.7) for 65-69 years, and 7.4 (5.5-9.9) for >70 years.

In this large study, elderly AA was associated with more severe thrombocytopenia, poorer response to first-line treatment, higher rate of dysplastic evolution and worse survival as compared to younger patients. This unmet need persists notwithstanding the availability of novel treatment modalities, highlighting the need for further optimization of clinical management.