Treatment of acquired aplastic anemia in children

Abstract

Aplastic anemia (AA) is an uncommon but serious disorder characterized by pancytopenia resulting from non-function of the bone marrow. The incidence of AA is approximately 3 fold more common in East Asia than in Europe and United States where yearly incidence rates are approximately two patients per one million.¹

Keywords:

• Aplastic anemia
• Children
• Immunosuppressive therapy
• Stem cell transplantation

Diagnosis of Childhood Aplastic Anemia

It is often difficult to distinguish hypoplastic myelodysplastic syndrome (MDS) from AA, especially in cases without cytogenetic abnormalities. Isolated anemia, which is the major presenting manifestation of refractory anemia (RA) in adults, is uncommon in children, who are more likely than adults to present with pancytopenia. In addition, hypocellularity of the bone marrow is more common in childhood RA. The new edition of the WHO classification for myeloid neoplasms outlines a provisional entity for refractory cytopenia for childhood (RCC) in which the diagnostic criteria for distinguishing RCC from AA are proposed.2

We studied the incidence and clinical feature of this new disease entity in Asian population with acquired bone marrow failure. A total of 100 children with cytopenia and hypocellular bone marrow (50 cases from Japan, 50 cases from China) were included in the present study. To obtain a diagnosis for cytopenia, at least two of the following must be present: (1) neutrophil count <1·5×10⁹/l; (2) hemoglobin <100 g/l; and (3) platelet count <50×10⁹/l. The severity of the disease was classified according to internationally accepted criteria.3 AA patients exhibited no morphological changes in any of their hematopoietic cell lineages. RCC was defined as persistent cytopenia with <5% blasts in the bone marrow and <2% blasts in the peripheral blood. In addition, RCC patients had <10% dysplastic changes in >2 cell lineages, or >10% in 1 cell lineage. Patients classified as having refractory cytopenia with multilineage dysplasia (RCMD) exhibited >10% of the dysplastic changes in >2 cell lineages. Dysplastic features of bone marrow aspirate cytology and trephine biopsy sampled were evaluated according to recommendations by the French–American–British Cooperative Leukemia Working Group and the morphology group of the European Working Group of MDS in Childhood.4,5 Final consensus for the diagnoses of 100 patients was follows: AA in 29 cases, RCC in 58 cases, and RCMD in 13 cases. The distribution of diagnoses were not different between Japanese cases and Chinese Cases; 12∶17 in AA; 33∶25 in RCC and 5∶8 in RCMD, respectively. Among the three groups there were no significant differences with regard to median age at diagnosis, sex, or days from onset of symptoms to diagnosis. While eight out of 29 (28%) patients in the AA group had very severe cytopenia, only two of the 58 patients (3%) in the RCC group and none of the 13 patients in the RCMD group had very severe cytopenia. On the other hand, 45 of the 58 patients (78%) in the RCC group and 10 of the 13 patients (77%) of the RCMD group had non-severe cytopenia (P<0·001). Additionally, 16 out of 29 AA patients (55%) exhibited severe hypoplastic bone marrow cellularity, while only four out of 58 RCC patients (7%) and none of the RCMD patients had severe hypoplastic bone marrow. A number of the RCC/RCMD patients exhibited mild to moderate hypocellularity (P<0·001). Data for cytogenetic analyses were available from 75 patients. Abnormal karyotypes were detected in one patient from the AA group (47,XX,+8) and in three patients in the RCC group (47,XX,+8; 46,XY, t(x∶3)(p11·2;q13); 49,idem,+6,+21).

The distribution of diagnoses in Japanese and Chinese children were 12∶17 for AA; 33∶25 for RCC and 5∶8 for RCMD, respectively. Patients with RCC/RCMD were milder in disease severity and bone marrow hypocellularity, compared to those with AA. According to the WHO classification system, it is recommended that children who satisfy the criteria for RCMD should be considered as RCC until the numbers of lineages involved are fully evaluated whether it is an important prognostic discriminator in childhood MDS.2 In the current study, 13 of the 71 MDS children (18%) were classified as RCMD. The bone marrow samples were more cellular, and dysplasia of cell morphology was more prominent in children with RCMD than those with RCC.

The most important aspect of the new proposal from the WHO classification system is whether the diagnosis has an impact on clinical outcomes including, response to treatment and incidence of late clonal diseases. Unfortunately, due to a short follow-up period and variety of treatments, we could not address this issue. To establish the new entity of RCC, future studies should prospectively compare the clinical outcomes between AA and RCC groups in a large number of patients.

First Line Treatment for Severe Aplastic Anemia in Children

Bone marrow transplantation (BMT) from an HLA-matched family donor (MFD) is the treatment of choice for severe aplastic anemia (SAA) in children. For children without an MFD, immunosuppressive therapy (IST) with a combination of antithymocyte globulin and cyclosporine has been successfully used.6 However, this decision is based on the results of comparative studies of these two therapies mainly conducted in the 1980s, and the outcome of both BMT and IST has dramatically improved over the past three decades.7,8 Therefore, updated evidences for treatment decision of pediatric SAA are required. We compared the outcome of children with SAA who received IST enrolled in the prospective multicenter trials of IST conducted by the Japan Childhood Aplastic Anemia Study Group9,10 or BMT from an MFD registered to the Transplant Registry Unified Management Program conducted by the Japan Society for Hematopoietic Cell Transplantation. The influence of potential risk factors on overall survival (OS) and failure free survival (FFS) was assessed according to first line treatment, time periods of treatment (1992–1999 and 2000–2009), age and other variables related to each treatment. FFS was defined as survival with treatment response. Death, primary or secondary graft failure, relapse, and second malignancy were considered treatment failure in patients who received BMT. Death, relapse, disease progression requiring stem cell transplantation from an alternative donor or 2nd IST, clonal evolution, and evolution to paroxysmal nocturnal hemoglobinuria were considered treatment failure in patients who received IST. Between 1992 and 2009, 599 children with SAA younger than 17 years received BMT from an MFD (n = 213) or IST (n = 386) as a first line treatment. While the OS did not differ between patients received IST and BMT (88±2 vs 90±2% at 15 years), the FFS was significantly inferior in patients received IST than in those received BMT (54±3 vs 84±3% at 15 years, P<0·0001). There was no significant improvement of outcomes in the two time periods; OS and FFS at 10 years in 1992–1999 vs 2000–2009 were 87±2 vs 93±2% and 66±3 vs 67±3%, respectively. In multivariate analysis, age <10 years was identified as a favorable factor for OS (P = 0·007) and choice of the first line IST was the only risk factor for FFS (P<0·0001). These updated data support the current algorism for treatment decision which recommends BMT when an MFD is available.6

HLA-mismatched Family Donors BMT

The first-line therapy for children with severe AA is allogeneic BMT from a human leukocyte antigen (HLA)-matched family donor, and IST is indicated for patients without HLA-matched family donors.6 While the standard therapy for children who fail to respond to IST is allogeneic BMT from an HLA-matched unrelated donor, BMT from an HLA-mismatched family donor has also been indicated. Compared with unrelated donors, an HLA-mismatched family donor has some advantages especially for children who need urgent transplantation.

We analyzed the clinical outcome of 578 children (325 boys and 253 girls) with AA (age, <20 years) who received allogeneic BMT between 1990 and 2009 in Japan, and registered to the Transplant Registry Unified Management Program. The median age at transplantation was 11 years (0–19), and the donors were serological 6/6 HLA-matched related donors (MRD) (n = 312), one locus-mismatched related donor (1MMRD) (n = 44), 2–3 loci-mismatched related (haploidentical) donors (n = 9), and HLA-matched unrelated donors (MUD) (n = 213). Causes of deaths were as follows: acute graft-versus-host disease (GVHD) (n = 4), chronic GVHD (n = 4), acute respiratory distress syndrome (n = 2), severe hemorrhage (n = 7), engraftment failure (n = 5), infection (n = 18), idiopathic pneumonia (n = 8), organ failure (n = 19), secondary malignancy (n = 4), and others (n = 4).

The probability of severe acute GVHD (grade III–IV) in patients transplanted from 1MMRD (26·9±7·4%) was significantly higher than that in patients transplanted from MRD (4·9±1·3%) (P<0·001). The probability of five-year overall survival (5y OS) of patients transplanted from 1MMRD (93·1±3·9%) was comparable to that of patients transplanted from MRD (93·1±1·5%) (P = not significant, NS), but it was significantly better than that of patients transplanted from haploidentical donors (66·7±15·7%, P = 0·016) and MUD (79·0±2·9%, P = 0·014). In the subgroup analysis of 1MMRD, no significant difference was observed between HLA class I-mismatched (n = 32) and class II-mismatched patients (n = 12) (5y OS; 93·8±4·3% vs 91·7±8·0%, P = NS). Comparison of the survival outcome based on the transplant period (1990–1999 vs 2000–2009) revealed that the probability of 5y OS of patients transplanted from 1MMRD was not significantly different [92·3±5·2% (n = 26) vs 94·4±5·4% (n = 18), P = NS], while that of patients transplanted from MUD significantly improved in the same period as we reported previously (67·1±5·5% (n = 73) vs 86·1±3·1% (n = 140), P = 0·001).11 In multivariate analysis, haploidentical donors (P<0·001), MUD (P<0·001), age ⩾10 years (P<0·001), and transplant period (1990–1999 vs 2000–2009, P = 0·006) were identified as independent covariates associated with unfavorable OS. Our analysis revealed that an HLA-mismatched related donor, especially 1MMRD, could be selected as a donor candidate for children with AA who need urgent transplantation.

Unrelated Donor BMT

Bone marrow transplantation from an unrelated donor (UBMT) is indicated as salvage therapy for patients with severe SAA who fail to respond to immunosuppressive therapy.12 In several recent studies, the effect of HLA high-resolution matching on outcome of patients who received a UBMT has been elucidated.13,14 However, results have been derived primarily from an analysis of patients with hematologic malignancies. Major obstacles for UBMT are different between patients with hematologic malignancies and patients with SAA. Relapse is a main cause of death for patients with hematologic malignancies, and Graft versus Leukemia effect may result in decrease in relapse rate. In contrast, graft failure is the main problem, and GVHD is the only negative effect for patients with SAA. Therefore, optimal HLA matching may be different between these two populations. Algorithms for donor selection derived from an analysis of patients with hematologic malignancies might not be useful for patients with SAA. However, a few studies have focused on the clinical significance of HLA-allele compatibility in patients with SAA.15 In a previous study, we analyzed the clinical significance of HLA-allele mismatching in 142 patients with SAA, in whom data of high-resolution typing of HLA-A, -B, and -DRB1 were available. Mismatching of HLA-A or -B alleles between donor and recipient was a strong risk factor for acute and chronic GVHD and OS, whereas mismatching of the HLA-DRB1 allele did not have a significant effect on patient outcomes.16 In the study from the National Marrow Donor Program, mismatching of HLA-DRB1 was the most crucial risk factor for OS.17 On the contrary, restricting BMT to donor-recipient pairs perfectly matched at high-resolution typing reduces the chance of undergoing UBMT for many patients. Therefore, strategies for selecting a partially HLA-allele mismatched donor are required when a full matched donor cannot be identified. We report a detailed analysis of outcome in 301 patients with SAA who were typed for HLA-A, -B, -C, -DRB1, -DQB1, and -DPB1 by a molecular technique and underwent UBMT through the JMDP.11 Of the 301 recipient/donor pairs, 101 (33·6%) were completely matched at 10 of 10 alleles, 69 (23%) were mismatched at one allele, and 131 (43·5%) were mismatched at ⩾2 alleles. Subjects were classified into five subgroups: complete match group (group I); single-allele mismatch group (groups II and III); multiple alleles restricted to HLA-C, -DRB1, and -DQB1 mismatch group (group IV); and others (group V). Multivariate analysis indicated that only HLA disparity of group V was a significant risk factor for poor survival and grade II–IV acute GVHD. HLA-DPB1 mismatching was not associated with any clinical outcome. We recommend the use of an HLA 10 of 10 allele-matched unrelated donor. However, if such a donor is not available, any single-allele or multiple-allele (HLA-C, -DRB1, and -DQB1) mismatched donor is acceptable as an unrelated donor for patients with severe AA.

Conclusion

Treatment for childhood SAA has advanced in the past decade, most notably with the improvement in survival after HLA-matched UBMT and HLA-mismatched family donor BMT. Creation of an SAA outcomes registry might contribute to the improvement of outcome in children with SAA.