Allogeneic hematopoietic stem cell transplantation (HSCT) is a curative therapy for sickle cell disease (SCD). In children, myeloablative HSCT from a human leukocyte antigen (HLA)-matched related donor (HLA-MRD) results in SCD-free survival rates of over 95%, with low rates of graft-versus-host disease (GVHD).¹ However, above 15 years transplantation is associated with a higher incidence of GVHD and transplant-related mortality.^{2, 3} SCD can be cured with mixed donor chimerism, making it possible to decrease conditioning intensity. The challenge is, therefore, ensuring long-term engraftment with sufficient donor chimerism to prevent the recurrence of SCD. The National Institutes of Health (NIH) has developed a non-myeloablative (NMA) conditioning regimen based on total-body irradiation (TBI) with 3 Gy associated with alemtuzumab.⁴ In adults, this approach leads to long-term mixed chimerism and immune tolerance, resulting in a 5-year overall survival of 93% without severe GVHD, but the incidence of graft failure, at 13%, may be higher than that after myeloablative transplantation.⁵

We report here the results of a study conducted by the Société Francophone de Greffe de Moelle et de Thérapie Cellulaire (SFGM-TC)comparing myeloablative and nonmyeloablative approaches in patients transplanted over the age of 15 years from a HLA-MRD.

SCD patients were eligible if they received a busulfan-based myeloablative conditioning (MAC) regimen with antithymocyte globulin (ATG), or a NMA regimen with 3 Gy TBI and alemtuzumab, according to the NIH protocol.⁴ We included 34 patients transplanted from January 2015 to October 2020: 20 in the MAC group and 14 in the NMA group. Median age at transplantation was 17 years (range: 15–46 years). The characteristics of the patients are shown in Table 1. The patients in the two groups were similar for all but two characteristics.: the patients in the MAC group were slightly younger (17 years [range: 15–36] vs. 18 years [range: 16–46] in the NMA group; p = .04), and the incidence of pretransplant cerebral vasculopathy was higher in the MAC group (70% vs. 6% in the NMA group; p < .001). The MAC regimen consisted of busulfan (12.8 mg/kg, intravenous), with cyclophosphamide (200 mg/kg) in 17 patients, and with fludarabine (150–160 mg/m²) in 3 patients. ATG (Thymoglobulin Sanofi-Genzyme) was added for all patients in the MAC group (20 mg/kg in 15 patients and 5–10 mg/kg in 5), to decrease graft failure and GVHD.¹ GVHD prophylaxis consisted of cyclosporine with methotrexate in 15 patients or with mycophenolate mofetil in 5. The stem cells source was bone marrow. In the NMA group, the conditioning regimen associated 3 Gy TBI and alemtuzumab (1 mg/kg), the stem cell source was G-CSF mobilized peripheral blood stem cells (PBSC) (at least 10 × 10⁶ CD34⁺ cells/kg recipient body weight, to improve engraftment) and posttransplant immunosuppression was achieved with sirolimus in all but one patient who received tacrolimus.

All patients engrafted, no secondary rejection was reported. Two-year overall survival and 2-year event-free survival (EFS), defined as the probability of being alive with whole-blood donor chimerism >20%, were similar in both groups: 95% (95% confidence interval [CI]: 86%–100%) in the MAC group (median follow-up of 51 months; range: 8–83) versus 100% (95% CI: 100%–100%) in the NMA group (median follow-up of 26 months; range: 16–47; p = .4). One patient who received MAC transplant died from GVHD and infection. These data compare favorably with those for previous series with more heterogeneous transplant modalities.³ We focused here on two conditioning regimens for which favorable outcomes have been reported in large series.^{1, 5}

Grade II–IV acute GVHD (aGVHD) was observed in four patients (20%) in the MAC group (Grade II in 3 and Grade III in 1) versus 0 in the NMA group (p = .12). Five (25%) patients in the MAC group developed chronic GVHD (cGVHD) (one mild and four moderate/severe), versus 0 in the NMA group (p = .046). All of the patients without GVHD were off immunosuppression 18 months after transplantation. Two-year EFS without Grade II–IV aGVHD or moderate/severe cGVHD (GEFS) was 70% (95% CI: 52%–93%) in the MAC group and 100% in the NMA group (p = .03) (Figure S1). None of the other variables tested—age, sex, cerebral vasculopathy, sex mismatch, ABO incompatibility, and CMV serology—was associated with GEFS. The incidence of cGVHD after MAC regimen in our series is similar to that reported for patients over the age of 15 years.¹ The absence of GVHD in the NMA group is perfectly consistent with published data, demonstrating an impressive efficacy of alemtuzumab and sirolimus for GVHD prophylaxis.^{4, 5}

One year after transplantation, median whole-blood donor chimerism was 97% (range: 58%–100%, N = 17) in the MAC group and 84% (range: 27%–97%, N = 14) in the NMA group (p = .022). More patients in the MAC group than in the NMA group experienced complete donor chimerism (Figure S2). However, donor chimerism remained above 20% in all patients, which is considered sufficiently high to reverse the SCD phenotype.^{1, 6} In the 29 patients with regular HbS determinations after transplantation, HbS levels were similar to those in the corresponding donor, except for the three patients with displayed donor chimerism rates below 50%. These three patients, all of whom were in the NMA group, had undergone transplantation from an AA donor. At their last follow-up (21, 44, and 20 months posttransplant) they had donor chimerism rates of 24%, 51%, and 40%, and HbS levels of 1.9%, 3%, and 26%, respectively. The patient with an HbS level of 26% presented signs of hemolytic anemia at the last follow-up. Hemolysis marker levels in all patients improved after transplantation, with no significant difference between the two groups (Table S1). None of the patients developed sickle cell-related complications after transplantation. Although most graft failures were reported to occur in the first 6 months posttransplant after the NIH protocol,⁴ we believe that longer follow-up is required to exclude long-term rejection in our study, particularly in patients with donor chimerism rates below 50%. One of the major issues with the NMA conditioning regimen is ensuring long-term engraftment with sufficient donor chimerism to be curative. It is, therefore, important to evaluate hemolysis parameters, late vaso-occlusive events and organ function very carefully.

Hematopoietic recovery was faster in the NMA group, with fewer platelet and red blood cell transfusions required (Table 1). In this group, neutrophil count did not fall below 0.5 × 10⁹/L in 20% of patients, and no platelet or red blood cell transfusions were required for 50% and 20% of patients, respectively. By contrast, in the MAC group, two patients needed a hematopoietic stem cell boost due to delayed engraftment. These data are consistent with the results of other series reporting faster hematopoietic recovery after the transplantation of PBSC than after bone marrow transplantation. Arterial hypertension, neurological complications, including seizures and posterior reversible encephalopathy syndrome (PRES), mucositis, and a need for enteral or parenteral nutrition were more frequently reported in the MAC group (Table 1). The cyclosporine used in the MAC group may explain the higher rates of hypertension, creatinine concentration increases and neurologic complications, such as PRES. There was no significant difference in the frequency of infectious complications of grade ≥3 between the two groups (Table S2). No CMV or posttransplant lymphoproliferative disease was reported. There was also no excess of viral infections in the NMA group, despite later CD4 and CD8 cell reconstitution (Figure S3). Finally, the median duration of hospitalization for transplantation was significantly shorter in the NMA group: 35 days (range: 21–52 days) versus 54 days (range: 39–192 days) for the MAC group (p < .001).

This study has several limitations. The study population was small despite the inclusion of all the patients from the SFGM-TC database who were transplanted with the two conditioning regimens chosen for comparison. We focused on the MAC regimen, based on busulfan and including ATG, due to our national extensive experience with this approach, but a treosulfan-based conditioning regimen might be a suitable alternative for reducing toxicity. The absence of graft failure in the NMA group might not be confirmed in a larger population with longer follow-up, potentially reaching an incidence of about 10%, as reported in published series.^{4, 5} Finally, longer follow-up is required especially after NMA regimen, to check that donor chimerism levels remain sufficiently high to prevent SCD, and to evaluate occurrence of myeloid malignancies, which have been reported after graft failure.

In conclusion, our data confirm the feasibility of matched related transplant in adolescents and young adults with SCD. Both busulfan-based MAC with ATG and NMA regimen with low-dose TBI and alemtuzumab can be used in these patients, resulting in excellent EFS. The NMA regimen is associated with lower early toxicity and lower cGVHD rates. Our data support the NMA approach, from adolescence onwards, and its consideration for children. However, the excellent engraftment after NMA regimen observed here requires confirmation in a larger population with longer follow-up. A prospective national trial, comparing NMA transplantation for adolescents and adults from a related HLA-matched donor with the standard of care for patients without HLA-identical donors is currently underway and may demonstrate long-term benefits of transplantation in adults with SCD (NCT04046705).

同种异体造血干细胞移植 (HSCT) 是一种治疗镰状细胞病 (SCD) 的疗法。在儿童中，来自人类白细胞抗原 (HLA) 匹配的相关供体 (HLA-MRD) 的清髓性 HSCT 导致无 SCD 存活率超过 95%，移植物抗宿主病 (GVHD) 发生率较低。¹然而，超过 15 年的移植与较高的 GVHD 发生率和移植相关死亡率相关。^{2, 3}SCD 可以通过混合供体嵌合体来治愈，从而可以降低调节强度。因此，挑战在于确保长期植入具有足够的供体嵌合体以防止 SCD 的复发。美国国立卫生研究院 (NIH) 开发了一种基于全身照射 (TBI) 的非清髓性 (NMA) 预处理方案，其中 3 Gy 与阿仑单抗相关。⁴在成人中，这种方法导致长期混合嵌合体和免疫耐受，导致 5 年总生存率为 93%，没有严重的 GVHD，但移植物失败的发生率为 13%，可能高于清髓后移植。^5个

我们在此报告由Société Francophone de Greffe de Moelle et de Thérapie Cellulaire (SFGM-TC)进行的一项研究的结果，该研究比较了 15 岁以上 HLA-MRD 移植患者的清髓性和非清髓性方法。

根据 NIH 方案，如果 SCD 患者接受了基于白消安的清髓性调节 (MAC) 方案和抗胸腺细胞球蛋白 (ATG)，或 NMA 方案和 3 Gy TBI 和阿仑单抗，则他们符合条件。⁴我们纳入了 2015 年 1 月至 2020 年 10 月移植的 34 名患者：MAC 组 20 名，NMA 组 14 名。移植时的中位年龄为 17 岁（范围：15-46 岁）。患者的特征如表 1 所示。两组患者除以下两个特征外均相似：MAC 组患者稍年轻（17 岁 [范围：15-36] 与 18 岁 [范围：16–46] 在 NMA 组；p  = .04），移植前脑血管病变的发生率在 MAC 组更高（70% 对 NMA 组 6%；p  <.001）。MAC 方案包括白消安（12.8 mg/kg，静脉注射），17 名患者使用环磷酰胺 (200 mg/kg)，3 名患者使用氟达拉滨 (150–160 mg/m 2 ) ^。为 MAC 组的所有患者添加 ATG（赛诺菲-健赞胸腺球蛋白）（15 名患者为 20 mg/kg，5 名患者为 5-10 mg/kg），以减少移植失败和 GVHD。¹ GVHD 预防包括 15 名患者使用环孢菌素和甲氨蝶呤或 5 名患者使用吗替麦考酚酯。干细胞来源是骨髓。NMA 组预处理方案联合 3 Gy TBI 和阿仑单抗（1 mg/kg），干细胞来源为 G-CSF 动员的外周血干细胞（PBSC）（至少 10 × 10 ⁶ CD34 ⁺细胞/千克受体体重，以改善植入）和西罗莫司在所有接受他克莫司治疗的患者中实现了移植后免疫抑制。

所有患者均植入，未报告二次排斥反应。两年总生存率和 2 年无事件生存率 (EFS)，定义为全血供体嵌合体存活的概率 >20%，两组相似：95%（95% 置信区间 [CI] MAC 组中的 86%–100%（中位随访 51 个月；范围：8–83）对比 NMA 组中的 100%（95% CI：100%–100%）（中位随访26 个月；范围：16–47；p  = .4)。一名接受 MAC 移植的患者死于 GVHD 和感染。这些数据与之前具有更多异质移植方式的系列数据相比具有优势。³我们在这里重点介绍了两种预处理方案，这些方案已在大型系列中报告了良好的结果。^{1, 5}

在 MAC 组的 4 名患者 (20%) 中观察到 II-IV 级急性 GVHD (aGVHD)（3 名 II 级和 1 名 III 级），而 NMA 组为 0 名（p = .12 ）  。MAC 组中有 5 名 (25%) 患者出现慢性 GVHD (cGVHD)（1 名轻度和 4 名中度/重度患者），而 NMA 组为 0 名 (p = .046 )  。所有没有 GVHD 的患者在移植后 18 个月都停止了免疫抑制。MAC 组无 II-IV 级 aGVHD 或中度/重度 cGVHD (GEFS) 的两年 EFS 为 70%（95% CI：52%-93%），NMA 组为 100%（p = .03)（图 S1）。测试的其他变量——年龄、性别、脑血管病变、性别不匹配、ABO 不相容和 CMV 血清学——均与 GEFS 相关。我们系列中 MAC 方案后 cGVHD 的发生率与 15 岁以上患者的报道相似。¹ NMA 组未发生 GVHD 与已发表的数据完全一致，证明阿仑单抗和西罗莫司对 GVHD 预防具有令人印象深刻的疗效。^{4, 5}

移植后一年， MAC 组的全血供体嵌合体中位数为 97%（范围：58%–100%， N = 17），NMA 组为 84%（范围：27%–97%， N  = 14）组 ( p  = .022)。与 NMA 组相比，MAC 组中更多的患者经历了完全供体嵌合现象（图 S2）。然而，供体嵌合体在所有患者中均保持在 20% 以上，这被认为足以逆转 SCD 表型。^{1, 6}在移植后定期进行 HbS 测定的 29 名患者中，HbS 水平与相应供体的 HbS 水平相似，除了显示的供体嵌合率低于 50% 的 3 名患者。这三名患者均属于 NMA 组，均接受了 AA 供体的移植。在最后一次随访时（移植后 21、44 和 20 个月），他们的供体嵌合率分别为 24%、51% 和 40%，HbS 水平分别为 1.9%、3% 和 26%。HbS 水平为 26% 的患者在最后一次随访时出现了溶血性贫血的迹象。所有患者的溶血标志物水平在移植后均有所改善，两组之间无显着差异（表 S1）。移植后没有患者出现镰状细胞相关并发症。⁴我们认为需要更长时间的随访才能排除我们研究中的长期排斥反应，尤其是供体嵌合率低于 50% 的患者。NMA 调理方案的主要问题之一是确保长期植入具有足够的供体嵌合体以进行治疗。因此，非常仔细地评估溶血参数、晚期血管闭塞事件和器官功能非常重要。

NMA 组的造血恢复更快，需要的血小板和红细胞输注更少（表 1）。在该组中，中性粒细胞计数未低于 0.5 × 10 ⁹/L 在 20% 的患者中，分别有 50% 和 20% 的患者不需要血小板或红细胞输注。相比之下，在 MAC 组中，两名患者由于移植延迟需要造血干细胞增强。这些数据与其他系列报道的 PBSC 移植后比骨髓移植后造血恢复更快的结果一致。MAC 组更常报告动脉高血压、神经系统并发症，包括癫痫发作和后部可逆性脑病综合征 (PRES)、粘膜炎以及需要肠内或肠外营养（表 1）。MAC 组中使用的环孢菌素可能解释了高血压、肌酐浓度升高和神经系统并发症（如 PRES）发生率较高的原因。两组间 ≥ 3 级感染并发症的发生率无显着差异（表 S2）。没有 CMV 或移植后淋巴增生性疾病的报道。尽管后来 CD4 和 CD8 细胞重建，但 NMA 组中也没有过量的病毒感染（图 S3）。最后，NMA 组移植住院的中位持续时间明显更短：35 天（范围：21-52 天）对比 MAC 组为 54 天（范围：39-192 天）（p  <.001）。

这项研究有几个局限性。尽管纳入了 SFGM-TC 数据库中的所有患者，这些患者接受了选择用于比较的两种预处理方案的移植，但研究人群仍然很小。我们专注于基于白消安并包括 ATG 的 MAC 方案，因为我们国家对这种方法有丰富的经验，但基于三硫丹的预处理方案可能是降低毒性的合适替代方案。NMA 组移植物失败的情况可能无法在更大的人群中得到证实，随访时间更长，可能达到约 10% 的发生率，如已发表的系列报道。^{4, 5}最后，需要更长时间的随访，尤其是在 NMA 方案之后，以检查供体嵌合体水平是否保持足够高以防止 SCD，并评估骨髓恶性肿瘤的发生，这些恶性肿瘤已在移植失败后报告。

总之，我们的数据证实了在患有 SCD 的青少年和年轻人中进行匹配相关移植的可行性。基于白消安的 MAC 联合 ATG 和 NMA 方案联合低剂量 TBI 和阿仑单抗均可用于这些患者，从而获得出色的 EFS。NMA 方案与较低的早期毒性和较低的 cGVHD 率相关。我们的数据支持从青春期开始的 NMA 方法及其对儿童的考虑。然而，这里观察到的 NMA 方案后的良好植入需要在更大的人群和更长时间的随访中得到证实。一项前瞻性的国家试验，