Bone marrow failure

Definition

• 1.

  Severe aplastic anemia (SAA) is defined by:

  • a.

    bone marrow cellularity of less than 25% and

  • b.

    at least two of the following cytopenias:

    • i.

      granulocyte count <500/µL (<200 µL defines very SAA),

    • ii.

      platelet count <20,000/µL, and/or

    • iii.

      reticulocyte count <20,000/µL.

Non-SAA occurs when the abovementioned criteria are not met. There is little consensus on distinguishing between mild and moderate aplastic anemia.

Pathophysiology

• Aplastic anemia results from an immunologically mediated, tissue-specific, organ-destructive mechanism. It is postulated that after exposure to an inciting antigen, cells and cytokines of the immune system destroy stem cells in the marrow resulting in pancytopenia. Treatment with immunosuppression can potentially lead to marrow recovery.

Gamma-interferon (γ-IFN) plays a central role in the pathophysiology of aplastic anemia. In vitro studies show that the T cells from patients with aplastic anemia secrete γ-IFN and tumor necrosis factor (TNF). Long-term bone marrow cultures have shown that γ-IFN and TNF are potent inhibitors of both early and late hematopoietic progenitor cells. Both of these cytokines suppress hematopoiesis by their effects on the mitotic cycle and, more importantly, by the mechanism of cell killing. The mechanism of cell killing involves the pathway of apoptosis (i.e., γ-IFN and TNF upregulate each other’s cellular receptors, as well as the Fas receptors in hematopoietic stem cells). Cytotoxic T cells also secrete interleukin-2 that causes polyclonal expansion of the T cells. Activation of the Fas receptor on the hematopoietic stem cell by the Fas ligand present on the lymphocytes leads to apoptosis of the targeted hematopoietic progenitor cells. Additionally, γ-IFN mediates its hematopoietic suppressive activity through IFN regulatory factor 1 that inhibits the transcription of cellular genes and their entry into the cell cycle. γ-IFN also induces the production of nitric oxide, the diffusion of which causes additional toxic effects on the hematopoietic progenitor cells. Direct cell–cell interactions between effective lymphocytes and targeted hematopoietic cells probably also occur. The oligoclonal expansion of CD4+ and CD8+ T cells that fluctuate with disease activity further supports an immune etiology.

Table 6.1 lists the various causes of acquired aplastic anemia.

Clinical manifestations

• Acquired aplastic anemia may be idiopathic or secondary. At least 70% of cases are idiopathic. The incidence is approximately two cases per million per year in the West and higher in parts of Asia (~4–7.5 cases per million per year) and the male:female ratio is 1:1. The onset of acquired aplastic anemia is usually gradual and the symptoms are related to the pancytopenia.

• Anemia results in pallor, easy fatigability, weakness, and loss of appetite.

• Thrombocytopenia leads to petechiae, easy bruising, severe nosebleeds, gastrointestinal bleeding, and hematuria.

• Leukopenia leads to increased susceptibility to infections and oral ulcerations and gingivitis that respond poorly to antibiotic therapy.

• Hepatosplenomegaly and lymphadenopathy do not generally occur and their presence may suggest an underlying malignant, rheumatologic, or metabolic process.

Laboratory investigations

• Anemia : normocytic or macrocytic, normochromic

• Reticulocytopenia : absolute count more reliable

• Leukopenia : granulocytopenia often less than 1500/µL

• Thrombocytopenia : platelets often less than 30,000/µL

• Fetal hemoglobin : may slightly to moderately elevated

• Bone marrow :

  • Marked depression or absence of hematopoietic cells and replacement by fatty tissue containing reticulum cells, lymphocytes, plasma cells, and usually tissue mast cells.

  • Megaloblastic changes and other features indicative of dyserythropoiesis frequently seen in the erythroid precursors.

  • Bone marrow biopsy essential to assess cellularity for diagnosis and to exclude the possibility of poor aspiration technique or poor bone marrow sampling; additionally, it will help to rule out granulomas, myelofibrosis, or leukemia.

  • Chromosomal analysis is normal and assists in excluding myelodysplastic syndromes (MDSs). Expert hematopathology assessment may help distinguish SAA from hypocellular MDS.

  • Bone marrow cultures, and molecular testing for infectious agents when indicated.

• Chromosome breakage assay : performed on peripheral blood to screen for FA

• Flow cytometry (CD59) : performed on peripheral blood to diagnose paroxysmal nocturnal hemoglobinuria (PNH)

• Telomere length : performed on peripheral blood to screen for DC

• Liver function chemistries : to exclude hepatitis

• Renal function chemistries : to exclude renal disease

• Viral serology testing : hepatitis A, B, and C antibody panel (although posthepatitis aplastic anemia is generally associated with seronegative hepatitis); Epstein–Barr virus (EBV) antibody panel; parvovirus B19 IgG and IgM antibodies; varicella antibody titer; cytomegalovirus (CMV) antibody titer; human immunodeficiency virus antibody test

• Quantitative immunoglobulins : to rule out immunodeficiency

• Autoimmune disease evaluation : antinuclear antibody, total hemolytic complement (CH50), C3, C4, and direct antiglobulin test

• Human leukocyte antigen ( HLA ) typing : patient and family, done at the diagnosis of SAA to identify a suitable donor and ensure a timely transplant

Physical examination, appropriate laboratory screening assays and imaging studies, and, if warranted, mutation analysis should be performed to rule out other IBMFSs (FA, DC, DBA, SDS, AMT as well as rare genetic causes).

Treatment

Table 6.4 shows the recommendations for the treatment of moderate and SAA.

Treatment choices and long-term follow-up

• Although the short-term outcome with IST is comparable to that obtained with HLA-matched related HSCT, the decision to choose HSCT for younger patients with a histocompatible donor is based on the result of long-term follow-up. There are low rates of late mortality (due to chronic GVHD and therapy-related cancer) in patients undergoing HSCT, and the survival curves are relatively flat. Improved GVHD prophylaxis and safer preparative regimens have further improved these results. In contrast, there is a high risk of clonal hematopoietic disorders (MDS, AML, and PNH) in patients treated with IST compared to HSCT. Patients undergoing IST must be closely followed for the development of clonal disorders. Given the risk of clonal evolution and the fact that outcomes for unrelated transplantation are similar to those seen with HLA-matched related donor transplants, there are active trials in children and adults to perform matched unrelated transplants at initial diagnosis with SAA.

Salvage therapy

• For patients who fail sibling donor HSCT, or have a partial response [absolute neutrophil count (ANC) ≥500/µL, but are red cell and platelet transfusion dependent], or relapse following IST, management choices include alternative donor HSCT or further IST. HSCT is preferred to IST if a suitable donor is available. Children and teenagers for whom a fully HLA-matched unrelated donor exists (as determined by high-resolution typing) are excellent candidates for an alternative donor HSCT. For patients without a good alternative donor, a second course of ATG/CSA is warranted although mismatched donors are being considered in some centers. Eltrombopag has been used for these patients in clinical trials.

Long-term sequelae and outcomes for SAA

• Outcomes for both IST and HSCT have improved considerably in recent years. The results of multiple cohorts report a slightly different response rate and incidence of the clonal evolution of SAA to MDS, AML, or PNH. These data vary based on length of follow-up, age of patient, as well as institution/consortium.

• Complete or partial response rates in the range of 60–70%, largely from studies in adults, have been reported with IST. Horse ATG appears to be superior to rabbit ATG in these studies. Although the outcomes in children for IST are generally superior to those described for adults, disease-free survival for matched related HSCT is ~95%.

• IST improves hematopoiesis and achieves transfusion independence in the majority of patients, but the time to response is long. Hematopoietic response may be partial and relapses are relatively common.

• The incidence of clonal hematopoietic disorders, including PNH, MDS, and AML in patients with SAA treated with IST, ranges from 10 to 40%. The European Bone Marrow Transplantation Working Party compared the rate of secondary malignancies following HSCT and IST. Forty-two malignancies developed in 860 patients receiving IST, compared to 9 in 748 patients who underwent HSCT. In this study, acute leukemia and MDS were seen exclusively in IST-treated patients while the incidence of solid tumors was similar in the two groups of patients.

• From the aggregate data, there are a number of conclusions:

  • Matched sibling donor HSCT is always superior as primary therapy in young patients (<20 years of age) at any neutrophil count.

  • IST, due to transplant-related morbidity and mortality in older patients, is superior to HSCT in older patients (41–50 years).

  • For the 21- to 40-year-old age-group the differences are less clear.

  • In all age-groups, there are a higher percentage of late failures and clonal evolution in the IST-treated patients.

  • When considering the response rate (partial and complete) for IST, the low rate of transplant failure with alternative donor transplant, the incidence of GVHD, and the evolution of clonal disease after IST, the difference in survival between patients treated with matched unrelated donor HSCT and IST increases with time. Thus matched unrelated transplant (preferably within controlled clinical trials) is being considered primary therapy for SAA.

Treatment of moderate aplastic anemia

• The natural history of moderate aplastic anemia is uncertain and clinical experience varies widely. For this reason, it is generally thought that these patients should be treated initially with supportive therapy with very close follow-up. The majority of patients progress to SAA or develop significant and severe thrombocytopenia and bleeding, serious infections, or a chronic red blood transfusion requirement. These patients should be treated with the same treatment options as described for SAA.

Pathophysiology and genetics

• 22 FA complementation groups have been thus far defined. All 22 FA genes have been cloned ( Table 6.2 ). Complementation groups FANCA, C, and G represent ~90% of the cases. The gene products of these genes have been shown to cooperate in a common pathway. After FANCM and the FA-associated protein FAAP24 detect DNA damage, eight of the FA proteins (FANCA, B, C, E, F, G, L, and M) assemble to form the FA core complex that is required to monoubiquitinate and activate FANCD2 and FANCI. Ubiquitinated FANCD2 and FANCI form a dimer that stabilizes the stalled replication fork and then in turn interacts in nuclear repair foci with the downstream FA gene products (FANCO, D1, N, and J) in the FA/BRCA DNA damage repair pathway. Damage repair is then achieved by the late FA proteins in cooperation with proteins from other DNA repair pathways. Of note, FANCD1 has been identified as BRCA2. Despite the identification of this pathway, the manner in which disruption in this cascade of events results in a faulty DNA damage response and genomic instability leading to hematopoietic failure, birth defects, and cancer predisposition is incompletely understood.

FA cells are characterized by hypersensitivity to chromosomal breakage as well as hypersensitivity to G2/M cell cycle arrest induced by DNA crosslinking agents. In addition, there is sensitivity to oxygen-free radicals and to ionizing radiation.

Clinical manifestations

• FA is inherited as an autosomal recessive disorder (>99%), rarely as an X-linked recessive (FANCB, <1%) or an autosomal dominant (FANCR/RAD51), and is the most frequently inherited aplastic anemia. FANCA is the most common complementation group, representing about 60–70% of cases. FANCC and FANCG are the next most common, representing ~10% of cases each. The other complementation groups are quite rare, representing the remainder of cases ( Table 6.2 ).

• Genotype–phenotype correlations are complex and are emerging and relate to the complementation group as well as the specific allelic mutation (i.e., null versus hypomorphic gene product). In particular, certain associations relating genotype to specific congenital anomalies, early-onset aplastic anemia, leukemia, as well as Wilms’ tumor and medulloblastoma, have been confirmed.

• All racial and ethnic groups are affected.

• Pancytopenia is the usual finding.

• The median age at hematologic presentation of patients with aplastic anemia is approximately 8–10 years. Leukemia tends to appear later in the teenage years and solid tumors appear in young adulthood and continue to occur as patients age.

• Hematologic dysfunction usually presents with macrocytosis, followed by thrombocytopenia, often leading to progressive pancytopenia and SAA. FA frequently terminates in MDS and/or AML.

• The diagnosis of FA should always be considered in any child with an isolated cytopenia even when the classical somatic anomalies are absent as a significant number of these cases are physically normal.

• FA cells are hypersensitive to chromosomal breaks induced by DNA crosslinking agents. This observation is the basis for the commonly used chromosome breakage test for FA. The clastogens diepoxybutane (DEB) and mitomycin C (MMC) are the agents most frequently used in vitro to induce chromosome breaks, gaps, rearrangements, quadriradii, and other structural abnormalities. Clastogens also induce cell cycle arrest in G2/M. The hypersensitivity of FA lymphocytes to G2/M arrest, detected using cell cycle analysis by flow cytometry either de novo or clastogen induced, is being used by some as a screening tool for FA.

• Bone marrow examination reveals hypocellularity and fatty replacement consistent with the degree of peripheral pancytopenia. Residual hematopoiesis may reveal dysplastic erythroid (megaloblastoid changes, multinuclearity) and myeloid (abnormal granulation) precursors and abnormal megakaryocytes.

• Congenital anomalies include increased pigmentation of the skin along with café-au-lait spots and hypopigmented areas, short stature (impaired growth hormone secretion), skeletal anomalies (especially involving the thumb, radius, and long bones), male hypogenitalism, microcephaly, abnormalities of the eyes (microphthalmia, strabismus, ptosis, and nystagmus) and ears (deafness), hyperreflexia, developmental delay, and renal and cardiac anomalies. However, up to 40% of patients lack obvious physical abnormalities. There is great clinical heterogeneity even within a genotype (affected siblings may be phenotypically different).

• There is a nearly 800-fold increased relative risk of developing AML and perhaps an even greater relative risk of nonhematologic solid tumors (e.g., squamous cell carcinoma of head and neck, cancer of the breast, kidney, lung, colon, bone, retinoblastoma, and female gynecologic) in patients with FA. In general, these occur at much younger ages than those seen in the general population. A relatively large number of patients only become aware that they have FA when they are diagnosed with cancer. Androgen-related liver neoplasia may also occur as adenoma or hepatoma and rarely carcinoma. The risk of solid tumors may become even higher as death from aplastic anemia is reduced, and as post-HSCT patients survive longer. These data must be considered in the context of HSCT, as the risk of nonhematologic malignancy is likely to increase as a result of HSCT conditioning regimens and chronic GVHD. Treatment for cancer is generally ineffective due to severe side effects of chemotherapy and radiation therapy experienced by these patients.

• Prenatal diagnosis is possible by amniotic fluid cell cultures and chorionic villus biopsy.