Yang Liu

← Primer

HSC · Healing Source Code

移植后第三天的那一针——半相合移植如何从最后的选择变成常规选择 The injection on day three — how haploidentical transplantation went from last resort to routine option

2026-07-25 · 半相合移植, 移植后环磷酰胺, GVHD, 造血干细胞移植

每个需要移植的病人身边都站着一群现成的供者:父母、子女、半数的兄弟姐妹——他们与你共享恰好一半的 HLA。血缘上这是最近的距离,免疫学上却曾是最远的。把这样一份骨髓输进去,等于同时点燃两场战争:供者的 T 细胞把受者当成异物,受者残余的免疫又要把它赶出去。半相合供者因此长期被写在候选名单的最后一行。把它挪上来的不是新细胞也不是新药,而是一个关于时机的判断:把一支用了半个世纪的老化疗药,挪到移植之后再打。

一堵按位点计价的墙

墙的高度很早就被量过。1990 年,西雅图的 Anasetti 等人在 Human Immunology 上分析了 281 例接受亲属骨髓移植的血液肿瘤患者,供者与患者共享一条 HLA 单倍型,在另一条上错配程度不等。结论近乎冷酷:每多错配一个位点,重度急性 GVHD 的相对风险乘以 1.95,错配越深生存越差。

本可以只用配型全合的供者——问题是很多人根本没有。2014 年,Gragert 等人在 N Engl J Med 上用美国骨髓库的数据算过:找到高分辨全合无关供者的概率,欧洲裔白人是 75%,南美/中美裔黑人只有 16%。越是祖源多样的人群越被挡在门外,而多数人都有一位半相合的亲属。

把士兵全部撤走

第一条路是釜底抽薪:闯祸的既然是供者 T 细胞,就把它们从移植物里拿掉。1998 年,佩鲁贾的 Aversa 等人在 N Engl J Med 上报告 43 例高危急性白血病患者,用强化 conditioning 加超大剂量、经 T cell depletion 的 CD34⁺ 细胞(部分病例另加去 T 骨髓),全部实现完全供者 engraftment,可评估者无一例发生急、慢性 GVHD——代价是 40% 的移植相关死亡率(原文未拆分死因)。士兵撤光,城池不再内乱,也就无人守卫。

一支迟到三天的化疗药

另一条线索埋在六十年代。1963 年 Berenbaum 与 Brown 在 Nature 上报告,给小鼠单剂 cyclophosphamide 可延长皮肤移植的存活;1966 年 Johns Hopkins 的 Santos 与 Owens 同样在 Nature 上证明,大鼠的 GVHD 可用细胞毒药物治疗。反直觉在于:cyclophosphamide 本是清髓用的重药,移植之后再给,反而保住了移植。不同细胞对它的耐受差得很远——1990 年 Kastan 等人在 Blood 上测出,人骨髓中造血祖细胞的 aldehyde dehydrogenase(ALDH,活化型 cyclophosphamide 的解毒酶)表达最高,淋巴细胞最低。刚植入的造血祖细胞自带解药,刚被激活的 T 细胞没有。

2001 年,Johns Hopkins 的 Luznik 等人在 Blood 上把它做成小鼠模型:非清髓 conditioning 加移植后第 3 天的 cyclophosphamide,MHC 不相合的细胞可以稳定长期植入。2008 年,他们在 Biology of Blood and Marrow Transplantation 上给出改变判断的人体数据:68 名患者接受非清髓 conditioning 与不去除 T 细胞的半相合骨髓移植,第 3 天(部分加第 4 天)给 50 mg/kg 的 cyclophosphamide,可评估者中 13% 移植物失败,II–IV 度急性 GVHD 34%、III–IV 度仅 6%,一年 nonrelapse mortality 15%。

机制被改写了一次

最初的解释很顺:cyclophosphamide 杀死正响应新抗原、快速增殖的 alloreactive T 细胞,而高 ALDH 的造血细胞和调节性 T 细胞(Treg)幸存。2013 年 Kanakry 等人在 Science Translational Medicine 上的人体数据支持了后半句:效应 Treg 迅速回升且高表达 ALDH,而在异种移植模型里去掉移植物中的 Treg,PTCy 的保护便随之消失。

改写发生在前半句。2019 年,NCI 的 Wachsmuth 等人在 J Clin Invest 上用未去除 T 细胞的小鼠半相合移植模型追踪 alloreactive T 细胞——它们并没有被清除,切除胸腺也不影响 PTCy 的效果。真正发生的是,这些细胞在功能上被打瘫:给药后一天之内就建立起足够强的抑制环境,而 Treg 的优先恢复分量随时间加重。Nunes 与 Kanakry 同年的综述给出临床旁证:PTCy 后有临床意义的急性 GVHD 依然常见,与”彻底删除”不相容。新证据多来自小鼠与体外体系,人体内的确切次序还在补。

从最后一行走进常规

裁决来得很快。BMT CTN 1101(NCT01597778)把 368 名化疗敏感的淋巴瘤或缓解期急性白血病患者在相同 conditioning 下随机分到双份脐带血或半相合骨髓。2021 年 Fuchs 等人在 Blood 上报告,主要终点两年无进展生存无统计学差异(35% 对 41%),但两年 nonrelapse mortality 18% 对 11%、总生存 46% 对 57%,次要终点(含总生存)偏向半相合。

它随后越过了半相合的边界。2023 年,BMT CTN 1703(NCT03959241)在 N Engl J Med 上给出三期结果:431 名接受减低强度 conditioning 的成人(供者为配型全合亲属、全合或 7/8 错配无关者),随机接受移植后第 3、4 天给药的 cyclophosphamide 加 tacrolimus 与 mycophenolate mofetil,或传统的 tacrolimus–methotrexate,一年无 GVHD 无复发生存 52.7% 对 34.9%(风险比 0.64),总生存、复发与 engraftment 未见明显差异。同年 Auletta 等人的 CIBMTR 登记数据显示,到 2020 年为止的五年里,美国半相合移植的使用随 PTCy 的普及显著上升,更依赖这些错配平台的正是族裔多样的患者——尽管最常用的供者仍是全合无关者。

账单的另一半

代价没有消失,只是换了科室。PTCy 之后,巨细胞病毒来得更多。2021 年 Goldsmith 等人在 Blood 上分析 CIBMTR 数据:180 天内巨细胞病毒感染的累积发生率,用 PTCy 的半相合组 42%、同胞全合组 37%,用钙调磷酸酶抑制剂的同胞全合组 23%——升高的是 PTCy 本身,与供者类型无关。

engraftment 同样不牢靠,非恶性疾病里尤其如此。2019 年 Bolaños-Meade 等人在 Lancet Haematology 上写道,他们此前用于重型血红蛋白病的半相合方案有一半患者移植物失败;把 total body irradiation 从 200 cGy 提到 400 cGy 后,17 例中只剩 1 例原发性移植物失败。

回到开头那群站在病床边的亲属。PTCy 证明的是,横在他们与患者之间的免疫屏障不必靠清除细胞才能跨越,它可以被药理学重新塑形:用一支便宜的老药,在正确的时间给下去。自体基因治疗则从根上取消了这道屏障:供者与受者是同一个人;但 conditioning 依然要腾出龛位,风险只是从”别人的免疫”挪到了”自己被改写的基因组”。被消除的从来不是代价,而是它出现的位置。

参考文献

  1. Anasetti C, Beatty PG, Storb R, et al. Effect of HLA incompatibility on graft-versus-host disease, relapse, and survival after marrow transplantation for patients with leukemia or lymphoma. Hum Immunol. 1990;29(2):79-91. DOI · PMID 2249952
  2. Gragert L, Eapen M, Williams E, et al. HLA match likelihoods for hematopoietic stem-cell grafts in the U.S. registry. N Engl J Med. 2014;371(4):339-48. DOI · PMID 25054717
  3. Aversa F, Tabilio A, Velardi A, et al. Treatment of high-risk acute leukemia with T-cell-depleted stem cells from related donors with one fully mismatched HLA haplotype. N Engl J Med. 1998;339(17):1186-93. DOI · PMID 9780338
  4. Berenbaum MC, Brown IN. Prolongation of homograft survival in mice with single doses of cyclophosphamide. Nature. 1963;200:84. DOI · PMID 14074645
  5. Santos GW, Owens AH. Production of graft-versus-host disease in the rat and its treatment with cytotoxic agents. Nature. 1966;210(5032):139-40. DOI · PMID 5962069
  6. Kastan MB, Schlaffer E, Russo JE, Colvin OM, Civin CI, Hilton J. Direct demonstration of elevated aldehyde dehydrogenase in human hematopoietic progenitor cells. Blood. 1990;75(10):1947-50. PMID 2337669
  7. Luznik L, Jalla S, Engstrom LW, Iannone R, Fuchs EJ. Durable engraftment of major histocompatibility complex-incompatible cells after nonmyeloablative conditioning with fludarabine, low-dose total body irradiation, and posttransplantation cyclophosphamide. Blood. 2001;98(12):3456-64. DOI · PMID 11719388
  8. Luznik L, O’Donnell PV, Symons HJ, et al. HLA-haploidentical bone marrow transplantation for hematologic malignancies using nonmyeloablative conditioning and high-dose, posttransplantation cyclophosphamide. Biol Blood Marrow Transplant. 2008;14(6):641-50. DOI · PMID 18489989
  9. Kanakry CG, Ganguly S, Zahurak M, et al. Aldehyde dehydrogenase expression drives human regulatory T cell resistance to posttransplantation cyclophosphamide. Sci Transl Med. 2013;5(211):211ra157. DOI · PMID 24225944
  10. Wachsmuth LP, Patterson MT, Eckhaus MA, Venzon DJ, Gress RE, Kanakry CG. Post-transplantation cyclophosphamide prevents graft-versus-host disease by inducing alloreactive T cell dysfunction and suppression. J Clin Invest. 2019;129(6):2357-2373. DOI · PMID 30913039
  11. Nunes NS, Kanakry CG. Mechanisms of graft-versus-host disease prevention by post-transplantation cyclophosphamide: an evolving understanding. Front Immunol. 2019;10:2668. DOI · PMID 31849930
  12. Fuchs EJ, O’Donnell PV, Eapen M, et al. Double unrelated umbilical cord blood vs HLA-haploidentical bone marrow transplantation: the BMT CTN 1101 trial. Blood. 2021;137(3):420-428. DOI · PMID 33475736 · NCT01597778
  13. Bolaños-Meade J, Hamadani M, Wu J, et al. Post-transplantation cyclophosphamide-based graft-versus-host disease prophylaxis. N Engl J Med. 2023;388(25):2338-2348. DOI · PMID 37342922 · NCT03959241
  14. Auletta JJ, Kou J, Chen M, et al. Real-world data showing trends and outcomes by race and ethnicity in allogeneic hematopoietic cell transplantation: a report from the Center for International Blood and Marrow Transplant Research. Transplant Cell Ther. 2023;29(6):346.e1-346.e10. DOI · PMID 36924931
  15. Goldsmith SR, Abid MB, Auletta JJ, et al. Posttransplant cyclophosphamide is associated with increased cytomegalovirus infection: a CIBMTR analysis. Blood. 2021;137(23):3291-3305. DOI · PMID 33657221
  16. Bolaños-Meade J, Cooke KR, Gamper CJ, et al. Effect of increased dose of total body irradiation on graft failure associated with HLA-haploidentical transplantation in patients with severe haemoglobinopathies: a prospective clinical trial. Lancet Haematol. 2019;6(4):e183-e193. DOI · PMID 30878319 · NCT00489281

Every patient who needs a transplant is surrounded by ready-made donors: parents, children, half of one’s siblings — people who share exactly one HLA haplotype. By blood this is the shortest distance; immunologically it was once the longest. Infusing such a marrow means lighting two wars at once: the donor’s T cells treat the recipient as foreign, and the recipient’s residual immunity tries to throw the graft back out. Haploidentical donors were therefore long written on the last line of the candidate list. What moved them up was neither a new cell nor a new drug, but a judgment about timing: taking a half-century-old chemotherapy agent and giving it after the transplant.

A wall priced by the locus

The height of the wall was measured early. In 1990, Anasetti and colleagues in Seattle analyzed, in Human Immunology, 281 patients with hematologic malignancies who received marrow from relatives; the donors shared one HLA haplotype with the patient and were mismatched to varying degrees on the other. The conclusion was almost cold: for each additional incompatible locus, the relative risk of severe acute GVHD was multiplied by 1.95, and the deeper the mismatch, the worse the survival.

One could simply use fully matched donors — except that many people do not have one. In 2014, Gragert and colleagues used data from the U.S. registry in N Engl J Med: the probability of finding a high-resolution fully matched unrelated donor was 75% among whites of European descent and only 16% among blacks of South or Central American descent. The more ancestrally diverse the population, the more it was shut out — while for most patients a relative mismatched at just a single haplotype is available.

Withdrawing every soldier

The first road was to remove the fuel: if the donor T cells are the culprits, take them out of the graft. In 1998, Aversa and colleagues in Perugia reported in N Engl J Med on 43 patients with high-risk acute leukemia treated with intensified conditioning plus very large doses of T cell–depleted CD34⁺ cells (with T cell–depleted bone marrow added in some cases); full donor-type engraftment was achieved in all of them, and no evaluable patient developed acute or chronic GVHD — at a cost of 40% transplantation-related mortality (the paper does not break down causes of death). With every soldier withdrawn, the city no longer fights itself, and no longer has defenders.

A chemotherapy drug that arrives three days late

The other clue was buried in the 1960s. In 1963 Berenbaum and Brown reported in Nature that a single dose of cyclophosphamide could prolong skin graft survival in mice; in 1966 Santos and Owens at Johns Hopkins showed, also in Nature, that GVHD in the rat could be treated with cytotoxic agents. The counterintuitive part: cyclophosphamide is a heavy drug used for myeloablation, yet given after the transplant it preserves the transplant instead. Different cells tolerate it very differently — in 1990 Kastan and colleagues measured in Blood that within human bone marrow it is the hematopoietic progenitors that carry the most aldehyde dehydrogenase (ALDH, the enzyme that detoxifies activated cyclophosphamide), and the lymphocytes the least. The freshly infused hematopoietic progenitors carry their own antidote; the freshly activated T cells do not.

In 2001, Luznik and colleagues at Johns Hopkins turned this into a mouse model in Blood: nonmyeloablative conditioning plus cyclophosphamide on day 3 after transplantation allowed MHC-incompatible cells to engraft stably over the long term. In 2008, in Biology of Blood and Marrow Transplantation, they produced the human data that changed judgments: 68 patients received nonmyeloablative conditioning and T cell–replete haploidentical bone marrow transplantation, with 50 mg/kg of cyclophosphamide on day 3 (in some, also on day 4); graft failure occurred in 13% of evaluable patients, grades II–IV acute GVHD in 34% and grades III–IV in only 6%, with 1-year nonrelapse mortality of 15%.

The mechanism was rewritten once

The original explanation ran smoothly: cyclophosphamide kills the alloreactive T cells that are responding to new antigens and proliferating fast, while hematopoietic cells and regulatory T cells (Treg) with high ALDH survive. Human data from Kanakry and colleagues in Science Translational Medicine in 2013 supported the second half: effector Tregs recovered rapidly and expressed relatively high levels of ALDH, and removing Tregs from the graft in a xenogeneic transplant model abolished the protection conferred by PTCy.

The rewriting happened in the first half. In 2019, Wachsmuth and colleagues at the NCI used a T cell–replete murine haploidentical transplantation model in J Clin Invest to track alloreactive T cells — they were not eliminated, and the thymus was not necessary for PTCy to work. What actually happened was that those cells were functionally impaired: a strongly suppressive environment was in place within a day of dosing, and the preferential recovery of Tregs mattered more and more as time went on. A review by Nunes and Kanakry the same year offered clinical corroboration: clinically significant acute GVHD occurs frequently after PTCy, which is inconsistent with outright elimination of alloreactive T cells. The new evidence comes largely from murine and in vitro systems; the exact sequence inside human beings is still being filled in.

From the last line into routine

The verdict came quickly. BMT CTN 1101 (NCT01597778) randomly assigned 368 patients with chemotherapy-sensitive lymphoma or acute leukemia in remission, under identical conditioning, to double umbilical cord blood or haploidentical bone marrow. In 2021 Fuchs and colleagues reported in Blood that the primary end point, 2-year progression-free survival, showed no statistically significant difference (35% versus 41%), but 2-year nonrelapse mortality was 18% versus 11% and overall survival 46% versus 57%; the secondary end points, overall survival among them, pointed toward haploidentical marrow.

It then crossed beyond the haploidentical boundary. In 2023, BMT CTN 1703 (NCT03959241) reported phase 3 results in N Engl J Med: 431 adults undergoing reduced-intensity conditioning (donors were HLA-matched relatives, or matched or 7/8 mismatched unrelated donors) were randomized to cyclophosphamide given on days 3 and 4 after transplantation plus tacrolimus and mycophenolate mofetil, or to conventional tacrolimus–methotrexate; 1-year GVHD-free, relapse-free survival was 52.7% versus 34.9% (hazard ratio 0.64), while overall survival, relapse and engraftment did not differ substantially. That same year, CIBMTR registry data from Auletta and colleagues showed that over the five years up to 2020 the use of haploidentical transplantation in the United States rose significantly alongside the spread of PTCy, and that it is ethnically diverse patients who most depend on these mismatched platforms — although matched unrelated donors remain the most common donor type.

The other half of the bill

The cost has not disappeared; it has only moved to another department. After PTCy, cytomegalovirus comes more often. In 2021 Goldsmith and colleagues analyzed CIBMTR data in Blood: cumulative incidences of CMV infection by day 180 were 42% in the haploidentical group given PTCy, 37% in the matched sibling group given PTCy, and 23% in the matched sibling group given a calcineurin inhibitor — it is PTCy itself, regardless of donor type, that is associated with the higher incidence.

Engraftment is not secure either, especially in non-malignant disease. In 2019 Bolaños-Meade and colleagues wrote in Lancet Haematology that their earlier haploidentical regimen for severe hemoglobinopathies had led to graft failure in half of the patients; after raising total body irradiation from 200 cGy to 400 cGy, only 1 of 17 patients had primary graft failure.

Return to those relatives standing beside the bed. What PTCy demonstrates is that the immune barrier between them and the patient need not be crossed by stripping cells away; it can be reshaped pharmacologically — with one cheap old drug, given at the right time. Autologous gene therapy removes the barrier at its root: donor and recipient are the same person. But conditioning still has to clear the niche, and the risk has merely moved from someone else’s immunity to one’s own rewritten genome. What gets abolished is never the cost, only the place where it appears.

References

  1. Anasetti C, Beatty PG, Storb R, et al. Effect of HLA incompatibility on graft-versus-host disease, relapse, and survival after marrow transplantation for patients with leukemia or lymphoma. Hum Immunol. 1990;29(2):79-91. DOI · PMID 2249952
  2. Gragert L, Eapen M, Williams E, et al. HLA match likelihoods for hematopoietic stem-cell grafts in the U.S. registry. N Engl J Med. 2014;371(4):339-48. DOI · PMID 25054717
  3. Aversa F, Tabilio A, Velardi A, et al. Treatment of high-risk acute leukemia with T-cell-depleted stem cells from related donors with one fully mismatched HLA haplotype. N Engl J Med. 1998;339(17):1186-93. DOI · PMID 9780338
  4. Berenbaum MC, Brown IN. Prolongation of homograft survival in mice with single doses of cyclophosphamide. Nature. 1963;200:84. DOI · PMID 14074645
  5. Santos GW, Owens AH. Production of graft-versus-host disease in the rat and its treatment with cytotoxic agents. Nature. 1966;210(5032):139-40. DOI · PMID 5962069
  6. Kastan MB, Schlaffer E, Russo JE, Colvin OM, Civin CI, Hilton J. Direct demonstration of elevated aldehyde dehydrogenase in human hematopoietic progenitor cells. Blood. 1990;75(10):1947-50. PMID 2337669
  7. Luznik L, Jalla S, Engstrom LW, Iannone R, Fuchs EJ. Durable engraftment of major histocompatibility complex-incompatible cells after nonmyeloablative conditioning with fludarabine, low-dose total body irradiation, and posttransplantation cyclophosphamide. Blood. 2001;98(12):3456-64. DOI · PMID 11719388
  8. Luznik L, O’Donnell PV, Symons HJ, et al. HLA-haploidentical bone marrow transplantation for hematologic malignancies using nonmyeloablative conditioning and high-dose, posttransplantation cyclophosphamide. Biol Blood Marrow Transplant. 2008;14(6):641-50. DOI · PMID 18489989
  9. Kanakry CG, Ganguly S, Zahurak M, et al. Aldehyde dehydrogenase expression drives human regulatory T cell resistance to posttransplantation cyclophosphamide. Sci Transl Med. 2013;5(211):211ra157. DOI · PMID 24225944
  10. Wachsmuth LP, Patterson MT, Eckhaus MA, Venzon DJ, Gress RE, Kanakry CG. Post-transplantation cyclophosphamide prevents graft-versus-host disease by inducing alloreactive T cell dysfunction and suppression. J Clin Invest. 2019;129(6):2357-2373. DOI · PMID 30913039
  11. Nunes NS, Kanakry CG. Mechanisms of graft-versus-host disease prevention by post-transplantation cyclophosphamide: an evolving understanding. Front Immunol. 2019;10:2668. DOI · PMID 31849930
  12. Fuchs EJ, O’Donnell PV, Eapen M, et al. Double unrelated umbilical cord blood vs HLA-haploidentical bone marrow transplantation: the BMT CTN 1101 trial. Blood. 2021;137(3):420-428. DOI · PMID 33475736 · NCT01597778
  13. Bolaños-Meade J, Hamadani M, Wu J, et al. Post-transplantation cyclophosphamide-based graft-versus-host disease prophylaxis. N Engl J Med. 2023;388(25):2338-2348. DOI · PMID 37342922 · NCT03959241
  14. Auletta JJ, Kou J, Chen M, et al. Real-world data showing trends and outcomes by race and ethnicity in allogeneic hematopoietic cell transplantation: a report from the Center for International Blood and Marrow Transplant Research. Transplant Cell Ther. 2023;29(6):346.e1-346.e10. DOI · PMID 36924931
  15. Goldsmith SR, Abid MB, Auletta JJ, et al. Posttransplant cyclophosphamide is associated with increased cytomegalovirus infection: a CIBMTR analysis. Blood. 2021;137(23):3291-3305. DOI · PMID 33657221
  16. Bolaños-Meade J, Cooke KR, Gamper CJ, et al. Effect of increased dose of total body irradiation on graft failure associated with HLA-haploidentical transplantation in patients with severe haemoglobinopathies: a prospective clinical trial. Lancet Haematol. 2019;6(4):e183-e193. DOI · PMID 30878319 · NCT00489281