Consensus recommendations for severe aplastic anemia

TO THE EDITOR:

In this issue of Blood Advances, a panel applied the Modified Delphi method to derive consensus recommendations for severe aplastic anemia (SAA).¹ Providing guidance for treating physicians has become important, with data extrapolated from published medical literature. These recommendations can be derived from synthesized and summarized data through systematic reviews, meta-analyses, and expert-based consensus statements using predefined methods such as the nominal group technique, Delphi, or RAND/UCLA method. Additionally, the GRADE (Grades of Recommendation, Assessment, Development, and Evaluation) framework offers a systematic approach to rating the quality of evidence and the strength of recommendations.

The GRADE framework sets the benchmark for assessing evidence quality and recommendation strength in health care. It mandates a detailed and systematic approach to scrutinizing study evidence, evaluating its quality, and formulating guidelines.² Several governmental organizations, agencies, and medical societies, including the American Society of Hematology, use GRADE to develop guidelines.

In the current consensus, the authors opted for a RAND/UCLA–modified Delphi method, which included 11 adult and pediatric experts. Although valuable, a Delphi consensus falls short of the rigorous standards required for guideline development. When high-quality evidence is lacking, recommendations are generally weak and/or associated with caveats. In these circumstances, expert recommendations can be complementary and often rely on what is adopted in their personal practices or respective institutions and not rooted in robust evidence supported by available methodologies.

The panel divided the recommendations into 4 categories: diagnosis, choice of initial therapy, supportive care, and subsequent management. For diagnosis, some recommendations lack robust evidence, such as using fluorescence in situ hybridization for myelodysplastic syndrome (MDS)–associated cytogenetics to exclude the diagnosis of hypocellular MDS and applying next-generation sequencing as standard practice for all cases at diagnosis. Although next-generation sequencing may identify MDS-defining variants that could inform prognosis and aid in follow-up for clonal evolution, its value in decision-making has not been firmly established, especially when used in isolation. Although intuitive, interpreting mutations at the diagnosis of SAA, in those unresponsive to Immunosuppressive Therapy (IST) and during follow-up after medical therapy also requires validation in prospective studies, and available data do not support performing these tests routinely due to the difficulty in interpreting the findings.^3,4 

There is growing interest in testing for germ line mutations in selected groups, as proposed by the panel, such as in those with age-adjusted short telomere lengths or suspected predisposition syndromes based on personal or family history or physical stigmata. Identifying these defects is typically associated with a lack of response to immunosuppressive therapy, a higher risk for hematopoietic stem cell transplantation (HSCT) complications, and influences donor selection, affecting decision-making.⁵ 

An area with wide variability and a paucity of data concerns prophylactic antimicrobial measures in SAA. In patients treated with lymphocyte-depleting therapy (Antithymocyte Globulin [ATG] or allogeneic HSCT), the panel recommends prophylaxis with antibacterials (including for pneumocystis pneumonia), antifungals (active against mold), and antivirals (for herpes simplex and varicella-zoster virus) until CD4 counts reach 200/μL and the absolute neutrophil count reaches ≥500/μL. It is worth noting that treatment algorithms in SAA include ATG or HSCT as initial therapy, depending on the age and availability of a histocompatible donor. Therefore, the prophylaxis recommendations herein apply to newly diagnosed patients with SAA. However, a lack of rigorous data specific to SAA supports this practice. Consequently, practice between centers is significantly variable, with preferences commonly extrapolated from other settings.

Regarding initial management, the panel has incorporated the paradigm applied for decades, which includes the age, severity of disease, and availability of a histocompatible donor primarily.⁶ Younger, fit patients aged ≤40 years with an HLA-matched sibling donor undergo a matched-related donor allogeneic HSCT. For those with a matched-unrelated donor, upfront HSCT can be considered. However, identifying and coordinating all the HSCT logistics in most centers worldwide could take several months, precluding this approach as standard in most places.

For patients without a histocompatible donor, those not eligible for HSCT, or those aged >40 years, the panel recommends horse ATG + cyclosporine for individuals aged ≤20 years and horse ATG + cyclosporine + eltrombopag (a 3-drug regimen) for those aged >20 years. Here, evidence supports the incorporation of eltrombopag to upfront standard IST in adults, whereas its benefit in children is unclear. In younger patients, the panel included a haploidentical HSCT upfront in younger fit patients (aged ≤20 years) after careful clinical consideration and stated that this modality is under investigation in those aged between 21 and 40 years. What characterized the careful consideration is left vague in the consensus. Haploidentical HSCT has the advantage of a readily available donor in most cases, with promising preliminary outcomes; however, the higher total body irradiation dose (400 cGy) required in the upfront setting to prevent rejection will likely lead to fertility issues and possibly other long-term consequences.⁷ In a large data set that included >1600 patients who underwent alternative donor HSCT for SAA, outcomes with haploidentical HSCT, when compared with histocompatible donor transplantation, did not fare as favorably.⁸ 

For those unresponsive to initial medical therapy or relapse (in particular early), the panel recommends an allogeneic HSCT with the best available donor in suitable patients, which has been increasingly accepted by the field. When an allogeneic HSCT is not an option, a repeat course of high-intensity immunosuppressive therapy (ATG based) is recommended for fitter patients, especially those with a late relapse. For those unfit for such intensive therapy, lower-intensity options without ATG are recommended, including cyclosporine, eltrombopag, androgens, growth factors, and transfusion support, which are reasonable approaches based on existing data. In the current consensus, the experts, all based in the United States, reiterated much of what has already been published in guidelines and expert reviews.^9-11 When data are limited or unavailable, experts' insights can be helpful, often reflecting their personal or institutional practices. A comprehensive and systematic approach to the literature using the GRADE methodology, which includes experts from around the globe, could reduce existing uncertainties, especially in rare diseases such as SAA. The availability of resources, economic factors, local policies, and patient preferences can significantly influence SAA management recommendations across different countries. This is particularly relevant, because SAA is more prevalent in some lower-middle income nations. Therefore, applying rigorous methodologies such as GRADE, with worldwide representation and unique regional perspectives, is welcomed and increasingly necessary.

Contribution: P.S. and A.G.K. wrote this article.

Conflict-of-interest disclosure: P.S. reports receiving consulting fees or honoraria for lectures, presentations, speakers’ bureaus, and advisory boards from AstraZeneca, Alexion, Bristol Myers Squibb (BMS), Janssen, Novartis, BioCryst, Pfizer, and Roche. A.G.K. reports receiving consulting fees or honoraria for lectures, presentations, speakers’ bureaus, and advisory boards from Agios, Alexion, Amgen, BioCryst, Celgene/BMS, Roche, Novartis, Pfizer, Samsung, and Sobi.

Correspondence: Phillip Scheinberg, Division of Hematology, Hospital A Beneficência Portuguesa, Rua Martiniano de Carvalho, 951, São Paulo, Brazil, 01321-001; email: scheinbp@bp.org.br.