How I treat acquired aplastic anemia

Abstract

Survival in severe aplastic anemia (SAA) has markedly improved in the past 4 decades because of advances in hematopoietic stem cell transplantation, immunosuppressive biologics and drugs, and supportive care. However, management of SAA patients remains challenging, both acutely in addressing the immediate consequences of pancytopenia and in the long term because of the disease's natural history and the consequences of therapy. Recent insights into pathophysiology have practical implications. We review key aspects of differential diagnosis, considerations in the choice of first- and second-line therapies, and the management of patients after immunosuppression, based on both a critical review of the recent literature and our large personal and research protocol experience of bone marrow failure in the Hematology Branch of the National Heart, Lung, and Blood Institute.

Figure 1

Algorithm for initial management of SAA. In patients who are not candidates for a matched related HSCT, immunosuppression with horse ATG plus cyclosporine should be the initial therapy. We assess for response at 3 and 6 months but usually wait 6 months before deciding on further interventions in case of nonresponders. In patients who are doing poorly clinically with persistent neutrophil count less than 200/μL, we proceed to salvage therapies earlier between 3 and 6 months. Transplant options are reassessed at 6 months, and donor availability, age, comorbidities, and neutrophil count become important considerations. We favor a matched unrelated HSCT in younger patients with a histocompatible donor and repeat immunosuppression for all other patients. In patients with a persistently low neutrophil count in the very severe range, we may consider a matched unrelated donor HSCT in older patients. In patients who remain refractory after 2 cycles of immunosuppression, further management is then individualized taking into consideration suitability for a higher risk HSCT (mismatched unrelated, haploidentical, or umbilical cord donor), age, comorbidities, neutrophil count, and overall clinical status. Some authorities in SAA consider 50 years of age as the cut-off for sibling HSCT as first-line therapy.

Figure 2

Long-term follow-up after immunosuppression. In patients treated with immunosuppression, we follow for relapse (among responders) and clonal evolution in all patients. A gradual downtrend in blood counts may signify hematologic response, underscoring the important of routine monitoring in this setting. In cases of relapse, we usually reintroduce more immunosuppression in the form of oral cyclosporine and/or a repeat course with rabbit ATG/CsA or alemtuzumab. In those who are unresponsive to more immunosuppression, further management will depend on suitability for HSCT (age, donor availability, comorbidities). When only higher risk HSCT options are available (mismatched unrelated, haploidentical, umbilical cord), we consider nonimmunosuppressive strategies, such as androgens (12-week trial), combination growth factors (G-CSF + Epo for 12 weeks), or experimental therapies. In patients with a very low neutrophil count unresponsive to G-CSF associated with infections, we consider a higher risk HSCT in younger patients. We monitor for clonal evolution by repeated marrow karyotype assessment at 6 and 12 months and then yearly thereafter. After 5 years, we tend to increase the interval between bone marrows. When faced with an abnormal karyotype, such as del13q, trisomy 6, pericentric inversion of chromosome 1;9, del20q, or trisomy 8, we assess for myelodysplasia by looking at blood counts, peripheral smear, and bone marrow morphology. On occasion, these karyotypes may not equate to progression to myelodysplasia and not be detected on repeated marrow examination. In cases where there is worsening blood counts and/or more significant dysplastic changes in the marrow, our approach is to seek transplant options, therapies for myelodysplasia, or a clinical trial. Monosomy 7 is almost never a transient finding and commonly associates to a more rapid progression to myelodysplasia and leukemia. In these cases, our approach is to seek HSCT earlier.

Figure 3

Durability of response after horse ATG. (A) Time to first late event among responders. The probability of a first late event (relapse or clonal evolution) among responders (N = 243) is approximately 50%. (B) In those who do not experience a late event, long-term survival in 10 years is excellent at 95%, whereas in those who experience a late event survival is not as favorable (65% in 10 years). (C) In our experience, high-risk evolution to monosomy 7, complex karyotype, high-grade myelodysplasia, or leukemia occurs in approximately 10% of responders long term. (D) Among responders who clonally evolved (any cytogenetic abnormality), survival was worse in those with a high-risk clonal event (monosomy 7, high-grade myelodysplasia, complex karyotype, or leukemia) compared with responders who do not experience high-risk evolution (principal karyotype findings in this lower risk group were trisomy 8 and del13q). Of note, among the high-risk clonal evolutions in responders, all occurred in those who achieved a partial hematologic response at 6 months after immunosuppression. (A,C) SD values (P = log-rank). Day 0 for all curves is the time of first horse ATG-based therapy. Data for other experimental immunosuppressive therapies as first-line are not shown. A late event is defined as either relapse or clonal evolution, whichever occurred first. Patients with repeated relapses or cytogenetic abnormalities were counted once at the time of first event.