背景：某些特殊类型的骨髓衰竭综合征，如肝炎相关性再生障碍性贫血（HAAA）、再生障碍性贫血（AA）后继发性骨髓增生异常综合征（MDS）以及分子国际预后评分系统（IPSS-M）高风险或输血依赖的低增生性骨髓增生异常肿瘤（MDS-h）/低原始细胞骨髓增生异常肿瘤（MDS-LB）患者均具有较高的死亡风险。目前，异基因造血干细胞移植（allo-HSCT）已成为血液病患者重要的治疗手段，但对于这些特殊亚型的疗效和安全性仍需进一步探索。

目的：评估allo-HSCT治疗特殊骨髓衰竭综合征的疗效和安全性，并探索优化移植预后的策略。

方法：回顾性分析2006年至2023年间本院接受allo-HSCT治疗的288例AA患者和392例MDS患者的临床资料。通过倾向性评分匹配（PSM）平衡混杂因素，对比30例HAAA患者和230例非HAAA患者的临床特征和移植预后；比较25例AA后继发性MDS患者、337例原发性MDS患者及285例未进展为MDS的AA患者的临床特征和移植预后；评估25例MDS-h和64例MDS-LB患者的移植指征及预后。重点分析HAAA患者移植后肝脏事件、AA后继发性MDS患者-7/del（7q）、MDS-h和MDS-LB患者动态输血依赖评估分别对三种疾病的具体影响，并按供者类型分层评估预后。

结果：平衡混杂因素后，HAAA患者接受allo-HSCT后的5年总生存（OS）率（75.8% vs 86.5%，P=0.409）、无失败生存（FFS）率（74.0% vs 83.2%，P=0.485）和无移植物抗宿主病无失败生存（GFFS）率（61.2% vs 67.6%，P=0.669）较非HAAA患者略低，但差异无统计学意义。两组患者在植入、感染以及移植物抗宿主病（GVHD）发生率方面无显著差异，免疫重建模式总体一致。与未出现移植后肝脏事件（如肝小静脉闭塞病、病毒性肝炎、肝脏GVHD以及ALT、AST或TBIL升高三倍）的HAAA患者相比，出现移植后肝脏事件的HAAA患者5年OS率显著降低，诊断至移植间隔≥75天可作为移植后肝脏事件的预测指标。

80%的AA后继发性MDS患者在MDS诊断前表现为非重型AA。与原发性MDS相比，继发性MDS患者的血小板计数较低，PNH克隆阳性率较高，低原始细胞患者（骨髓＜5%且外周血＜2%）占比较高。遗传学分析显示：继发性MDS患者GATA2和PIGA突变较常见，而原发性MDS患者U2AF1突变较常见；继发性MDS患者常表现为-7/del（7q）独立出现，-7/del（7q）增加了继发性MDS患者的移植复发风险，但并未影响移植后生存率。尽管继发性MDS患者的植入速度更慢，但其总体预后与无MDS进展的AA患者相似；此外，继发性MDS患者接受allo-HSCT后的3年OS率（85.9% vs 66.2%，P=0.079）、3年FFS率（79.2% vs 58.6%，P=0.110）以及1年移植相关死亡率的累积发生率（8.0% vs 28.1%，P=0.066），略优于原发性MDS患者，接近统计学意义。

25例MDS-h和64例MDS-LB患者，中位移植年龄43岁。常见基因突变包括U2AF1（32.4%）、ASXL1（24.3%）和RUNX1（20.3%）。动态输血依赖评估显示，患者输血依赖发生率随时间推移而增加。86.52%的MDS-h和MDS-LB患者因IPSS-M高危或输血依赖接受了allo-HSCT，移植后脱离输血依赖比例达83.1%。IPSS-M风险等级与诊断时输血依赖发生率相关（36.4% vs 71.2%，P=0.011），但诊断时或移植时输血依赖及IPSS-M高风险并不影响移植后生存情况。移植后3年OS和FFS率分别为81.7%和75.5%，1年复发累积发生率仅为1.2%，优化预处理方案或可提高移植后获益。

本研究按供者类型分别对三种亚型患者进行预后分层。对于HAAA患者，单倍体相合供者（HID）组患者移植后5年OS、FFS和GFFS率略低于同胞全相合供者（MSD）组，但差异无统计学意义；HID组患者移植后早期巨细胞病毒（CMV）血症发生率更高，CMV疾病的发生率为5.6%。对于AA后继发性MDS患者，采用HID或MSD移植后预后无显著差异。对于MDS-h和MDS-LB患者，与MSD组相比，HID组患者移植后急性GVHD累积发生率略高，但未达到显著性，两组患者移植后生存率、复发及慢性GVHD累积发生率均相似。

结论：HAAA患者、AA后继发性MDS、MDS-h和MDS-LB患者均具有独特的临床特征。校正潜在混杂因素后，HAAA患者的移植疗效与非HAAA患者相当；AA后继发性MDS移植后生存率略优于原发性MDS患者，接近统计学差异，且与未进展为MDS的AA患者相似；MDS-h和MDS-LB患者移植后输血依赖显著改善，IPSS-M高风险患者移植预后良好。以上结果进一步证实了allo-HSCT治疗这些特殊类型骨髓衰竭综合征的安全性和有效性。

Background: Certain rare subtypes of bone marrow failure syndromes, including hepatitis-associated aplastic anemia (HAAA), secondary myelodysplastic syndrome (MDS) following aplastic anemia (AA), and hypoplastic myelodysplastic neoplasm (MDS-h) or MDS with low blasts (MDS-LB) with higher-risk Molecular International Prognostic Scoring System (IPSS-M) or transfusion dependence, are associated with high mortality. Despite the importance of allogeneic hematopoietic stem cell transplantation (allo-HSCT), its efficacy and safety in these specific subtypes require further investigation.

Objective: To investigate the efficacy and safety of allo-HSCT in special bone marrow failure syndromes and to explore strategies for optimizing post-transplant outcomes.

Methods: A retrospective analysis was conducted on the clinical data of 288 patients with AA and 392 patients with MDS who underwent allo-HSCT at our institution from 2006 to 2023. Propensity score matching (PSM) was used to balance confounding factors. We compared clinical characteristics and transplant outcomes between 30 HAAA and 230 non-HAAA patients, between 25 AA secondary MDS, 337 de novo MDS, and 285 AA patients without MDS progression, and evaluated the transplant indications and outcomes of 25 MDS-h and 64 MDS-LB patients. The study focuses on analyzing the specific impact of post-transplant hepatic events in HAAA patients, -7/del(7q) in AA-derived secondary MDS patients, and the dynamic assessment of transfusion dependence in MDS-h and MDS-LB patients. Additionally, prognostic evaluation is stratified based on donor type.

Results: After adjusting for confounders, HAAA patients had slightly lower 5-year overall survival (OS) (75.8% vs 86.5%, P=0.409), failure-free survival (FFS) (74.0% vs 83.2%, P=0.485), and graft-versus-host disease (GVHD)-free failure-free survival (GFFS) (61.2% vs 67.6%, P=0.669) compared to non-HAAA patients, though differences were not statistically significant. No significant differences were observed in engraftment, infection, or GVHD incidence between the two groups, and overall immune reconstitution patterns were consistent. Compared with HAAA patients who did not develop post-transplant hepatic events—such as hepatic veno-occlusive disease, viral hepatitis, hepatic GVHD, or a threefold elevation in ALT, AST, or TBIL—those who experienced such events had a significantly lower 5-year OS rate. Additionally, a diagnosis-to-transplant interval of ≥75 days was identified as a predictor of post-transplant hepatic events.

Eighty percent of patients with secondary MDS following AA initially presented as non-severe AA before the diagnosis of MDS. Compared to de novo MDS, secondary MDS was characterized by lower platelet counts, a higher frequency of PNH clone positivity, and a greater proportion of low-blast patients (bone marrow blasts ＜5% and peripheral blasts ＜2%). Genetic analysis revealed that GATA2 and PIGA mutations were more frequently observed in secondary MDS, whereas U2AF1 mutations were more common in de novo MDS. Additionally, secondary MDS was often associated with isolated -7/del(7q), which increased the risk of post-transplant relapse but did not compromise post-transplant survival. Although engraftment was slower in secondary MDS, overall prognosis was comparable to that of AA patients without MDS progression. Furthermore, secondary MDS patients demonstrated marginally better outcomes than de novo MDS patients, with a 3-year OS rate of 85.9% vs 66.2% (P=0.079), a 3-year FFS rate of 79.2% vs 58.6% (P=0.110), and a 1-year cumulative incidence of transplant-related mortality of 8.0% vs 28.1% (P=0.066), approaching statistical significance.

For MDS-h and MDS-LB patients (n=89), the median age at transplant was 43 years. Common gene mutations included U2AF1 (32.4%), ASXL1 (24.3%), and RUNX1 (20.3%). Dynamic transfusion dependence assessment showed an increasing rate of transfusion dependence over time. 86.52% of MDS-h and MDS-LB patients underwent allo-HSCT due to higher-risk IPSS-M or transfusion dependence, with 83.1% achieving transfusion independence post-transplant. IPSS-M risk category was associated with transfusion dependence at diagnosis (36.4% vs 71.2%, P=0.011), but neither transfusion dependence at diagnosis or transplant nor higher-risk IPSS-M affected post-transplant survival. The 3-year OS and FFS rates were 81.7% and 75.5%, respectively, with a 1-year cumulative incidence of relapse of only 1.2%. Optimizing conditioning regimens may further enhance post-transplant benefits.

This study stratified the prognosis of the three patient subtypes based on donor type. In HAAA patients, the 5-year OS, FFS, and GFFS rates were slightly lower in the haploidentical donor (HID) group compared to the matched sibling donor (MSD) group, though the differences were not statistically significant. The HID group exhibited a higher incidence of early post-transplant cytomegalovirus (CMV) viremia, with a CMV disease incidence of 5.6%. In patients with secondary MDS following AA, no significant differences in prognosis were observed between those receiving HID or MSD transplantation. For patients with MDS-h and MDS-LB, the cumulative incidence of acute GVHD was slightly higher in the HID group compared to the MSD group, but the difference was not statistically significant. Post-transplant survival, relapse rates, and the cumulative incidence of chronic GVHD were comparable between the two groups.

Conclusions: Patients with HAAA, secondary MDS following AA, MDS-h, and MDS-LB exhibit distinct clinical characteristics. After adjusting for potential confounding factors, post-transplant outcomes in HAAA patients were comparable to those in non-HAAA patients. Survival after transplantation in secondary MDS following AA was marginally better than in de novo MDS, approaching statistical significance, and was similar to that of AA patients without MDS progression. In MDS-h and MDS-LB patients, transfusion dependence significantly improves after transplantation, while IPSS-M higher-risk patients demonstrate favorable transplant outcomes. These findings further support the safety and efficacy of allo-HSCT in the treatment of these rare subtypes of bone marrow failure syndromes.