Yang Liu

← Primer

HSC · Healing Source Code

Zynteglo 与 Skysona:两条 lentiviral 长期安全数据线 Zynteglo and Skysona: two long-term lentiviral safety records

2026-10-09 · lentiviral vector, insertional oncogenesis, beti-cel, eli-cel, Zynteglo, Skysona, β-地中海贫血, CALD

2022 年夏末,美国 FDA 先后批准了两款 lentiviral vector 的自体造血干细胞基因疗法:8 月 17 日是用于需要规律输血的 β-地中海贫血的 Zynteglo,9 月 16 日是用于早期活动性脑型肾上腺脑白质营养不良(CALD)的 Skysona。它们出自同一家公司,走同一条技术路线:取出患者自己的 CD34⁺ 细胞,在体外用病毒载体整合一个正常基因,清髓后回输。几年之后,Zynteglo 的长期随访汇总里没有恶性肿瘤的报告,而 Skysona 的说明书写明,血液肿瘤已在约 15% 的试验参与者中出现。这篇整理这两条数据线,以及它们为什么不能简单地并排比较。

两个先到的成功

lentiviral 基因添加在这两种病上的成功来得都比较早。2009 年,Cartier 等人在 Science 上报告了两名没有配型供者的 ALD 男孩:自体 CD34⁺ 细胞经携带野生型 ABCD1 的 lentiviral vector 修正后回输,出现多克隆重建,从输注后 14 到 16 个月起,进行性的脑脱髓鞘停止了,临床结果与异基因移植相当。

β-地中海贫血的路径更曲折,因为矫正后的造血干细胞没有选择性优势,而红系需要大量 β-globin。2010 年,Cavazzana-Calvo 等人在 Nature 上报告了第一位患者:一名 βE/β0 型成人,基因转移 33 个月后摆脱输血已 21 个月。他的大部分疗效来自一个偏髓系的优势克隆,载体插入导致 HMGA2 在红系细胞中被转录激活;作者的判断是,这种与疗效相伴的克隆优势可能只是巧合、随机,也可能是 HMGA2 失调带来的一种迄今良性的扩增。

随后是 BB305 载体(beti-cel,商品名 Zynteglo)的系统试验。2018 年,Thompson 等人在 N Engl J Med 上报告了两项 phase 1-2 研究(HGB-204,NCT01745120;HGB-205,NCT02151526):22 名患者中,13 名 non-β0/β0 基因型者除 1 人外都停止了输血,未观察到与载体整合相关的 clonal dominance。2022 年 Locatelli 等人报告的 phase 3 研究(HGB-207,NCT02906202)中,20/22(91%)可评估患者达到 transfusion independence,未观察到癌症;2024 年 Kwiatkowski 等人在 Lancet 上报告的 HGB-212(NCT03207009)纳入 β0/β0 等更严重基因型,18 名患者中 16 名(89%)达到并维持 transfusion independence。

eli-cel(Skysona)一侧,2017 年 Eichler 等人在 N Engl J Med 上报告了 STARBEAM 研究(ALD-102,NCT01896102)的中期结果:17 名早期 CALD 男孩,中位随访 29.4 个月,15 名(88%)存活且无重大功能残疾,当时未见载体偏向整合在已知原癌基因附近,也未见 clonal outgrowth。2024 年的更新为 32 名患者,中位随访 6 年,26 名(81%)没有重大功能残疾。

迟来的肿瘤

转折在随访里出现。2024 年,Duncan 等人在 N Engl J Med 上整理了 eli-cel 的临床试验数据(ALD-102、ALD-104 NCT03852498 及长期随访 LTF-304 NCT02698579):67 名患者中有 7 人发生血液肿瘤,其中 ALD-102 的 32 人中 1 例,ALD-104 的 35 人中 6 例,包括 MDS 和 1 例 AML,诊断时间从治疗后 14 个月到 92 个月不等。能分析的 6 例里,优势克隆都含有多个位点的载体插入,5 例涉及 MECOM-EVI1,1 例涉及 PRDM16;7 例中 6 例还带有 KRAS、NRAS、WT1、CDKN2A/B 或 RUNX1 的体细胞突变,另 1 例有 7 号染色体单体。论文的结论是,这些病例与原癌基因内的克隆性载体插入以及体细胞遗传缺陷的克隆演化相关。

FDA 于 2024 年 11 月发出安全通报,2025 年 8 月 7 日批准了要求的说明书修改。通报写道,Skysona 2022 年获批时,67 名临床研究患者中有 3 人(4%)报告了 MDS;截至 2025 年 7 月,10 人(15%)确诊血液肿瘤,诊断时间为用药后 14 个月到 10 年,其中 9 人接受了异基因移植,1 人死于与肿瘤治疗相关的原因,另有患者在 Skysona 来得及起效之前就已发病。说明书(boxed warning 保留)现在写明:所有接受 Skysona 治疗的临床研究患者都有 MECOM 内的整合,但尚不清楚哪些整合会导致恶性变;要求至少每 3 个月做一次血常规,并在第一年至少两次、此后每年评估克隆扩增。适应证也被收窄:Skysona 现在只用于没有 HLA 匹配异基因供者的 4 至 17 岁早期活动性 CALD 男孩。

相比之下,Zynteglo 的长期数据很安静。2026 年 Kwiatkowski 等人在 Blood 上报告了 LTF-303 的汇总:63 名受试者,中位随访 5.9 年,phase 3 参与者中 37/41(90.2%)达到 transfusion independence 并持续至末次随访,汇总中未报告恶性肿瘤、insertional oncogenesis,也未报告载体来源的 replication-competent lentivirus。但”未报告”不等于风险为零:Zynteglo 的说明书仍然列有 insertional oncogenesis 的潜在风险,并要求至少 15 年的监测。同一家公司的 Lyfgenia(镰状细胞病)说明书则带有血液肿瘤的 boxed warning。

能比,与不能比

把这两条数据线并排放,需要先说清楚它们不是同一类证据:疾病、患者年龄与基线不同,载体不同,受试者数量(63 与 67)和随访深度也不同,本文引用的研究也没有把两者放在同一个框架里比较。因此,“Zynteglo 没有肿瘤”不能读成”Zynteglo 更安全”,只能读成”在目前报告的随访里没有观察到”。

说明书里有两处白纸黑字的差别:Zynteglo 的 BB305 载体用细胞来源、红系特异的 promoter;Skysona 的 Lenti-D 载体用内部 MNDU3 promoter,说明书称这是一个经过改造的病毒 promoter,能在造血干细胞及其所有谱系中驱动转基因表达。两者都被说明书描述为 self-inactivating 载体。另一个事实同样来自说明书:所有 Skysona 患者都有 MECOM 内的整合,而多数人没有发病,所以单有整合不足以解释肿瘤。至于风险的差异有多少来自 promoter 设计,有多少来自疾病背景、预处理、转导细胞的负荷或其他因素,现有文献并未把这些变量拆开,所以 promoter 差异只能算一个尚未检验的假说,不是已知的原因。

还没说完的部分

Zynteglo 和 Skysona 的两条数据线放在一起,说明的是 lentiviral 基因添加的基因毒性不是”有或没有”的二元问题:它取决于疾病与载体,也需要以年为单位的随访才能看清,eli-cel 的肿瘤最晚在用药后 10 年才被确诊。两款疗法的长期随访都按设计持续约 15 年(Zynteglo 的 LTF-303 计划总计 15 年,Skysona 有 15 年的上市后随访要求)。而异基因移植那条老路的边界,即需要匹配供者、有排异风险,恰好决定了今天 Skysona 还能留给谁。

公司层面的变化也值得一句:Zynteglo 2019 年 5 月 29 日在欧盟获得有条件上市许可,2022 年 3 月 24 日应上市许可持有人 bluebird bio 的要求被撤销,EMA 记录的原因是该公司出于商业原因永久停止销售。bluebird 2025 年 2 月宣布将被 Carlyle 与 SK Capital 收购,6 月 2 日完成交易,9 月更名为 Genetix Biotherapeutics;该公司 2026 年 3 月的新闻稿称,2025 年在美国治疗了超过 100 名患者,并出现了公司首个盈利季度。

参考文献

  1. Cartier N, Hacein-Bey-Abina S, Bartholomae CC, et al. Hematopoietic stem cell gene therapy with a lentiviral vector in X-linked adrenoleukodystrophy. Science. 2009;326(5954):818-23. DOI
  2. Cavazzana-Calvo M, Payen E, Negre O, et al. Transfusion independence and HMGA2 activation after gene therapy of human β-thalassaemia. Nature. 2010;467(7313):318-22. DOI
  3. Eichler F, Duncan C, Musolino PL, et al. Hematopoietic stem-cell gene therapy for cerebral adrenoleukodystrophy. N Engl J Med. 2017;377(17):1630-1638. DOI
  4. Thompson AA, Walters MC, Kwiatkowski J, et al. Gene therapy in patients with transfusion-dependent β-thalassemia. N Engl J Med. 2018;378(16):1479-1493. DOI
  5. Locatelli F, Thompson AA, Kwiatkowski JL, et al. Betibeglogene autotemcel gene therapy for non-β0/β0 genotype β-thalassemia. N Engl J Med. 2022;386(5):415-427. DOI
  6. Eichler FS, Duncan CN, Musolino PL, et al. Lentiviral gene therapy for cerebral adrenoleukodystrophy. N Engl J Med. 2024;391(14):1302-1312. DOI
  7. Duncan CN, Bledsoe JR, Grzywacz B, et al. Hematologic cancer after gene therapy for cerebral adrenoleukodystrophy. N Engl J Med. 2024;391(14):1287-1301. DOI
  8. Kwiatkowski JL, Walters MC, Hongeng S, et al. Betibeglogene autotemcel gene therapy in patients with transfusion-dependent, severe genotype β-thalassaemia (HGB-212): a non-randomised, multicentre, single-arm, open-label, single-dose, phase 3 trial. Lancet. 2024;404(10468):2175-2186. DOI
  9. Kwiatkowski JL, Thompson AA, Schneiderman J, et al. Long-term efficacy and safety results of betibeglogene autotemcel gene therapy for transfusion-dependent β-thalassemia. Blood. 2026;147(19):2203-2214. DOI
  10. U.S. FDA. SKYSONA (elivaldogene autotemcel) prescribing information, revised 7/2025. 链接
  11. U.S. FDA. FDA approves required labeling changes for increased risk of hematologic malignancy following treatment with Skysona (elivaldogene autotemcel). 2025-08-07. 链接
  12. U.S. FDA. ZYNTEGLO (betibeglogene autotemcel) prescribing information. 链接
  13. U.S. FDA. LYFGENIA (lovotibeglogene autotemcel) prescribing information. 链接
  14. FDA news release (reproduced via PR Newswire). FDA approves first cell-based gene therapy to treat adult and pediatric patients with beta-thalassemia who require regular blood transfusions (2022-08-17). 链接
  15. European Medicines Agency. Public statement: Zynteglo, withdrawal of the marketing authorisation in the European Union. 2022-03-30. 链接
  16. Genetix Biotherapeutics (formerly bluebird bio). Press releases of 2025-09-18 and 2026-03-02. rebrand / 2025 results
  17. bluebird bio. Announces Definitive Agreement to be Acquired by Carlyle and SK Capital (2025-02-21); Announces Completion of Acquisition by Carlyle and SK Capital (2025-06-02, SEC Form 8-K Exhibit 99.1). 完成公告

In late summer 2022, the US FDA approved two autologous hematopoietic stem cell gene therapies built on lentiviral vectors, one after the other: on August 17, Zynteglo, for β-thalassemia requiring regular transfusions, and on September 16, Skysona, for early active cerebral adrenoleukodystrophy (CALD). They came from the same company and followed the same technical route: take the patient’s own CD34⁺ cells, use a viral vector to integrate a normal gene ex vivo, then reinfuse after myeloablation. A few years on, the pooled long-term follow-up of Zynteglo contains no reports of malignancy, while the Skysona label states that hematologic cancers have appeared in about 15% of trial participants. This piece goes through the two records, and through why they cannot simply be set side by side.

Two early successes

Lentiviral gene addition succeeded early in both diseases. In 2009, Cartier and colleagues reported in Science on two boys with ALD who had no matched donor: their autologous CD34⁺ cells were corrected with a lentiviral vector carrying wild-type ABCD1 and reinfused, polyclonal reconstitution followed, and from 14 to 16 months after infusion the progressive cerebral demyelination stopped, with clinical outcomes comparable to allogeneic transplantation.

The path in β-thalassemia was more winding, because corrected hematopoietic stem cells have no selective advantage and the erythroid lineage needs large amounts of β-globin. In 2010, Cavazzana-Calvo and colleagues reported the first patient in Nature: an adult with βE/β0 disease who had been transfusion-free for 21 months at 33 months after gene transfer. Most of his benefit came from a dominant, myeloid-biased clone in which vector insertion had led to transcriptional activation of HMGA2 in erythroid cells; the authors’ reading was that this clonal dominance accompanying the benefit might be simply coincidental and random, or might reflect a so-far benign expansion caused by HMGA2 dysregulation.

Next came the systematic trials of the BB305 vector (beti-cel, trade name Zynteglo). In 2018, Thompson and colleagues reported two phase 1-2 studies in N Engl J Med (HGB-204, NCT01745120; HGB-205, NCT02151526): of 22 patients, the 13 with a non-β0/β0 genotype all stopped transfusions except one, and no clonal dominance related to vector integration was observed. In the phase 3 study reported by Locatelli and colleagues in 2022 (HGB-207, NCT02906202), 20/22 (91%) evaluable patients reached transfusion independence and no cancers were observed; HGB-212 (NCT03207009), reported by Kwiatkowski and colleagues in Lancet in 2024, enrolled more severe genotypes including β0/β0, and 16 of 18 patients (89%) achieved and maintained transfusion independence.

On the eli-cel (Skysona) side, in 2017 Eichler and colleagues reported interim results of the STARBEAM study (ALD-102, NCT01896102) in N Engl J Med: of 17 boys with early CALD, followed for a median of 29.4 months, 15 (88%) were alive without major functional disability, and at that point no preferential vector integration near known proto-oncogenes and no clonal outgrowth had been seen. The 2024 update covered 32 patients followed for a median of 6 years, of whom 26 (81%) had no major functional disability.

The tumors that came late

The turn came in the follow-up. In 2024, Duncan and colleagues compiled the eli-cel clinical trial data in N Engl J Med (ALD-102, ALD-104 NCT03852498, and the long-term follow-up LTF-304 NCT02698579): 7 of 67 patients developed hematologic cancers, 1 of the 32 in ALD-102 and 6 of the 35 in ALD-104, including MDS and one case of AML, with diagnoses ranging from 14 to 92 months after treatment. In the 6 cases that could be analyzed, the dominant clones all carried vector insertions at multiple sites, with MECOM-EVI1 involved in 5 and PRDM16 in 1; 6 of the 7 also carried somatic mutations in KRAS, NRAS, WT1, CDKN2A/B or RUNX1, and the other had monosomy 7. The paper concluded that these cases were associated with clonal vector insertion within proto-oncogenes and with clonal evolution of somatic genetic defects.

FDA issued a safety communication in November 2024 and approved the required labeling changes on August 7, 2025. The communication states that when Skysona was approved in 2022, 3 of the 67 clinical study patients (4%) had reported MDS; by July 2025, 10 (15%) had a confirmed hematologic cancer, diagnosed from 14 months to 10 years after treatment, of whom 9 underwent allogeneic transplantation and 1 died of causes related to cancer treatment, and another patient had fallen ill before Skysona had had time to take effect. The label (the boxed warning is retained) now states that all clinical study patients treated with Skysona had integrations within MECOM, but it is not known which integrations lead to malignant transformation; it requires a complete blood count at least every 3 months, and an assessment for clonal expansion at least twice in the first year and annually thereafter. The indication was also narrowed: Skysona is now used only in boys aged 4 to 17 years with early active CALD who do not have an HLA-matched allogeneic donor.

By comparison, the long-term data for Zynteglo are quiet. In 2026, Kwiatkowski and colleagues reported the LTF-303 pooled analysis in Blood: 63 subjects, median follow-up 5.9 years, with 37/41 (90.2%) of phase 3 participants reaching transfusion independence that continued to the last follow-up, and no reports of malignancy, insertional oncogenesis, or vector-derived replication-competent lentivirus in the pooled analysis. But not reported does not mean zero risk: the Zynteglo label still lists a potential risk of insertional oncogenesis and calls for at least 15 years of monitoring. The label of Lyfgenia (for sickle cell disease), from the same company, carries a boxed warning for hematologic cancer.

What can and cannot be compared

Before putting these two records side by side, it has to be clear that they are not the same kind of evidence: the diseases, patient ages and baselines differ, the vectors differ, the number of subjects (63 versus 67) and the depth of follow-up differ, and none of the studies cited here compares the two within a single framework. So the absence of tumors with Zynteglo cannot be read as Zynteglo being safer; it can only be read as nothing having been observed in the follow-up reported so far.

The labels contain two plainly stated differences: Zynteglo’s BB305 vector uses a cell-derived, erythroid-specific promoter, while Skysona’s Lenti-D vector uses an internal MNDU3 promoter, which the label describes as a modified viral promoter that drives transgene expression in hematopoietic stem cells and all of their lineages. Both are described in the labels as self-inactivating vectors. Another fact also comes from the label: every Skysona patient had integrations within MECOM, yet most did not develop cancer, so integration alone does not explain the tumors. How much of the difference in risk comes from promoter design, and how much from disease background, conditioning, the burden of transduced cells or other factors, the existing literature has not separated; the promoter difference therefore remains an untested hypothesis, not a known cause.

What is still unsaid

Taken together, the two records show that the genotoxicity of lentiviral gene addition is not a binary of present or absent: it depends on the disease and the vector, and it takes follow-up measured in years to see clearly, since the latest eli-cel tumor was diagnosed 10 years after treatment. Long-term follow-up for both therapies is designed to last about 15 years (LTF-303 for Zynteglo is planned for 15 years in total, and Skysona carries a 15-year post-marketing follow-up requirement). And the limits of the older route, allogeneic transplantation, which requires a matched donor and carries the risk of rejection, are exactly what now determines whom Skysona is still left for.

The company side deserves a sentence too: Zynteglo received conditional marketing authorisation in the EU on May 29, 2019, and on March 24, 2022 it was withdrawn at the request of the marketing authorisation holder, bluebird bio; the reason recorded by the EMA was that the company had permanently discontinued marketing for commercial reasons. bluebird announced in February 2025 that it would be acquired by Carlyle and SK Capital, completed the transaction on June 2, and in September was renamed Genetix Biotherapeutics; the company’s March 2026 press release states that it treated more than 100 patients in the US in 2025 and had its first profitable quarter.

References

  1. Cartier N, Hacein-Bey-Abina S, Bartholomae CC, et al. Hematopoietic stem cell gene therapy with a lentiviral vector in X-linked adrenoleukodystrophy. Science. 2009;326(5954):818-23. DOI
  2. Cavazzana-Calvo M, Payen E, Negre O, et al. Transfusion independence and HMGA2 activation after gene therapy of human β-thalassaemia. Nature. 2010;467(7313):318-22. DOI
  3. Eichler F, Duncan C, Musolino PL, et al. Hematopoietic stem-cell gene therapy for cerebral adrenoleukodystrophy. N Engl J Med. 2017;377(17):1630-1638. DOI
  4. Thompson AA, Walters MC, Kwiatkowski J, et al. Gene therapy in patients with transfusion-dependent β-thalassemia. N Engl J Med. 2018;378(16):1479-1493. DOI
  5. Locatelli F, Thompson AA, Kwiatkowski JL, et al. Betibeglogene autotemcel gene therapy for non-β0/β0 genotype β-thalassemia. N Engl J Med. 2022;386(5):415-427. DOI
  6. Eichler FS, Duncan CN, Musolino PL, et al. Lentiviral gene therapy for cerebral adrenoleukodystrophy. N Engl J Med. 2024;391(14):1302-1312. DOI
  7. Duncan CN, Bledsoe JR, Grzywacz B, et al. Hematologic cancer after gene therapy for cerebral adrenoleukodystrophy. N Engl J Med. 2024;391(14):1287-1301. DOI
  8. Kwiatkowski JL, Walters MC, Hongeng S, et al. Betibeglogene autotemcel gene therapy in patients with transfusion-dependent, severe genotype β-thalassaemia (HGB-212): a non-randomised, multicentre, single-arm, open-label, single-dose, phase 3 trial. Lancet. 2024;404(10468):2175-2186. DOI
  9. Kwiatkowski JL, Thompson AA, Schneiderman J, et al. Long-term efficacy and safety results of betibeglogene autotemcel gene therapy for transfusion-dependent β-thalassemia. Blood. 2026;147(19):2203-2214. DOI
  10. U.S. FDA. SKYSONA (elivaldogene autotemcel) prescribing information, revised 7/2025. link
  11. U.S. FDA. FDA approves required labeling changes for increased risk of hematologic malignancy following treatment with Skysona (elivaldogene autotemcel). 2025-08-07. link
  12. U.S. FDA. ZYNTEGLO (betibeglogene autotemcel) prescribing information. link
  13. U.S. FDA. LYFGENIA (lovotibeglogene autotemcel) prescribing information. link
  14. FDA news release (reproduced via PR Newswire). FDA approves first cell-based gene therapy to treat adult and pediatric patients with beta-thalassemia who require regular blood transfusions (2022-08-17). link
  15. European Medicines Agency. Public statement: Zynteglo, withdrawal of the marketing authorisation in the European Union. 2022-03-30. link
  16. Genetix Biotherapeutics (formerly bluebird bio). Press releases of 2025-09-18 and 2026-03-02. rebrand / 2025 results
  17. bluebird bio. Announces Definitive Agreement to be Acquired by Carlyle and SK Capital (2025-02-21); Announces Completion of Acquisition by Carlyle and SK Capital (2025-06-02, SEC Form 8-K Exhibit 99.1). completion announcement