Pediatric Aplastic Anemia and Refractory Cytopenia of Childhood: A Retrospective Single Institutional Analysis to Determine Outcomes and Histomorphological Predictors

Introduction

Pediatric acquired aplastic anemia (AA) is a hypocellular bone marrow condition that is often difficult to distinguish from inherited bone marrow failure syndromes (IBMFS) and hypoplastic refractory cytopenia of childhood (RCC). RCC is a provisional entity for which the clinical implications are still under investigation. Historically, patients with RCC have been classified as AA. In this study, we sought to assess the intra-observer reproducibility and prognostic value of the histological criteria for RCC and AA as defined in the WHO classification. Specifically, we evaluated if RCC is an independent prognostic factor of overall survival (OS) and event-free survival (EFS), including treatment failure, disease progression and clonal evolution.

Methods

We performed a retrospective analysis of 149 AA patients seen at a single center between 1976-2010. Patients that met clinical or molecular diagnostic criteria for inherited bone marrow failure syndromes were excluded. We evaluated 5-year EFS, OS, and response rate to immunosuppressive therapy (IST); outcomes after matched HLA-identical related donor transplant (MRD) and matched unrelated donor (MUD) stem cell transplantation, as well as toxicities and clonal evolution. Of the 149 eligible individuals, 72 had diagnostic pathology material available for review. A double-blinded analysis of bone marrow aspirate and biopsy slides was undertaken by 3 pediatric hematopathologists. Each pathologist reviewed the slides of the entire study set and independently made a determination of the diagnosis. In cases where one or more differed in their assessment, the three conferred with an additional senior pediatric hematopathologist and all four reached agreement on a ‘consensus diagnosis’. To describe the degree of concordance between the assessments of the pathologists, a kappa coefficient was calculated. A logistic regression model was used to identify factors prognostic of treatment failure. Survival curves were generated using the methods of Kaplan and Meier.

Results

The 149 patients were classified as moderate aplastic anemia (MAA) (n=58), severe aplastic anemia (SAA) (n=50) and very severe aplastic anemia (vSAA) (n=41). Ninety-one patients received IST, 50 underwent a MRD HSCT, and 8 were observed without treatment. The concordance between pathologists in the assessment of the diagnosis of AA or RCC was modest, but ultimately a consensus between the pathologists was reached. The overall EFS and OS were 50.8% and 73.1% for all patients. The 5-year OS and 5-year EFS for all patients receiving IST were 87.8% and 51.5%, respectively, with a failure to respond to IST in 48%. Patients with vSAA had worse 5-year OS compared to SAA and MAA patients, respectively (p=0.02). 5-year OS was comparable at 83.6%±7.2 and 85.3%±5.2 for patients receiving a MUD or MRD respectively. Within AA patients receiving IST, none of the following factors had prognostic value to predict failure of treatment: consensus diagnosis of RCC, macrocytosis (MCV > 100), or HgF% > 4%. An MCV >100and HgF% >4% likewise were not associated with consensus diagnosis of RCC. However, a significantly higher hemoglobin (Hgb) and absolute reticulocyte count was found in patients with RCC compared to AA. Surprisingly the diagnosis of RCC was associated with a trend towards improved 5- year OS (84.9% versus 72.5%) and 5-year EFS (71.3% versus 52.5%) compared to AA. Five patients (~3%) of the analytical cohort all with the clinical diagnosis of MAA experienced clonal evolution or disease progression. Interestingly, the clinical diagnosis of vSAA significantly correlated with the histologic appearance of AA, whereas SAA and MAA corresponded with RCC, strongly suggesting a correlation between lower disease severity as determined by peripheral counts and/or marrow cellularity and the diagnosis of RCC.

Conclusion

The EFS and OS of the IST group in this single institution study is consistent with larger series. Our data indicate a low rate of clonal evolution in pediatric AA that in this series was associated exclusively with MAA. Most important, our findings suggest that the histologic diagnosis of RCC, reached by consensus among four pediatric hematopathologists does not predict disease outcome in this retrospective series of AA. These data indicate the need for larger prospective studies to determine the clinical significance of the classification.