Graft versus host disease (GVHD) is a major complication in human leukocyte antigen (HLA) matched allogeneic hematopoietic stem cell transplantation (allo-HSCT). Pharmacologic GVHD prophylaxis typically include calcineurin inhibitors, mycophenolate mofetil (MMF), methotrexate, and sirolimus. Other methods for GVHD prophylaxis include T-cell depletion, however, these approaches could increase the risk of infections and relapse, and potentially have variable results of delayed immune reconstitution.^1-3

Two doses of post-transplant cyclophosphamide (PTCy) have shown efficacy as GVHD prophylaxis in haploidentical (haplo)-HSCT. Extrapolating from its success in haplo-HSCT, PTCy has also demonstrated its effectiveness as a single agent in myeloablative HLA-matched-related and unrelated bone marrow transplant,^{2, 3} and in combination with other immunosuppressive agents in reduced-intensity peripheral blood stem cell (PBSC) transplant.^{4, 5} Furthermore, the combination of PTCy with other agents has retrospectively been found to be effective in myeloablative PBSC transplants.¹ However, initial studies with two doses of PTCy, in the reduced intensity setting, raised concern for increased risk of relapse.⁶ Furthermore, PTCy may increase the risk of infectious complictions, particularly cytomegalovirus (CMV) infection.⁷ In the effort of balancing between GVHD and graft versus tumor effect, one dose of PTCy has been tested. In non-myeloablative haploidentical bone marrow transplant, there was a trend towards higher extensive chronic GVHD (cGVHD) with one dose, when compared to two doses. There was no significant difference in the incidence of infection.⁸ Whether one dose of PTCy is sufficient for GVHD prophylaxis in myeloablative, matched-unrelated PBSC transplant, remains unclear.

We conducted a single-arm phase II clinical trial of one dose of PTCy, tacrolimus, and MMF for GVHD prophylaxis following myeloablative, matched-unrelated, PBSC allo-HSCT. The trial was registered with clinicaltrials.gov (NCT02065154). The primary endpoint of the study was the cumulative incidence (CI) of grade II-IV acute GVHD (aGVHD) at Day 100 post-transplant. Secondary endpoints included cumulative incidence of cGVHD, overall survival (OS), progression-free survival (PFS), relapse, and non-relapse related mortality (NRM). Details regarding eligibility, preparative regimens, GVHD prophylaxis, diagnosis and treatment, supportive care and statistical analysis are provided in Appendix S1.

Thirty-nine patients were enrolled between September 2013 and June 2018. Demographic and clinical characteristics are detailed in Table S1. Thirty-six donors were 8/8 HLA matched unrelated, and three were 7/8 matched (A mismatch = 1, B mismatch = 1, C mismatch = 1) unrelated. The trial was later amended to exclude patients with 7/8 match with evolving data showing more effective GVHD control by using the traditional two doses of PTCy.

There was one death, prior to engraftment, within the first 30 days. The remaining 38 (97%) patients successfully engrafted. The median time to ANC and platelet engraftment was 12 days (range 9–14 days) and 33 days (range 28–54 days), respectively. The 1-year data for donor chimerism and immune reconstitution are provided in Figures S1A, S1B, respectively.

The CI of grades II-IV and III-IV aGVHD at Day 100 post-transplant was 30% (95% CI 15.2%–40.1%) (Figure 1A) and 5% (95% CI 2.1%–8.2%) (Figure 1B), respectively (Table S2). There were two patients with grade IV aGVHD (lower gastrointestinal tract involvement) and none with grade III. The 2-year CI of mild, moderate, and severe cGVHD was 15% (95% CI 7.1%–23.3%), 15% (95% CI 8.4%–23.3%) and 18% (95% CI 12.6%–27.4%), respectively. For patients with cGVHD, skin was the most common organ involved (n = 9). Six patients had liver GVHD (Bu/Flu = 6), whereas four had lung involvement (Bu/Flu =3, Cy/TBI =1).

There was no significant difference in the occurrence of acute and chronic GVHD for age at diagnosis (≥60 years vs. <60 years and ≥55 years vs. <55 years), gender, conditioning regimen (Bu/Flu vs. others, TBI vs. non-TBI), disease (AML vs. others), DRI (high vs. others) and HLA-typing (8/8 vs. 7/8).

The 2-year CI of relapse and NRM was 21% (95% CI 7.6%–33.4%) and 34% (95% CI 18.6%–49.6%), respectively (Figure S2). There were four non-relapse related deaths prior to Day 100 post-transplant. Three of those patients were >60 years old. One patient died before neutrophil engraftment. The cause of death in the remaining three patients was sepsis (Enterococcus faecalis = 1, Staphylococcus epidermidis = 1, organism unidentified = 1), leading to multi-organ failure. For patients expiring beyond Day 100 post-transplant, the most common causes of death were cGVHD (13%) followed by infection without underlying GVHD (5%). In patients dying from cGVHD, the leading secondary causes of death included respiratory failure due to underlying lung GVHD and infection.

The 2-year estimate of PFS and OS was 45% (95% CI 29.5%–61.3%) and 51% (95% CI 34.7%–66.5%), respectively. The median OS was 36 months (95% CI 0–71.8 months) with a median follow up of 50 months (range 6–71 months) (Figure S3).

Three patients had 7/8 matched-unrelated donors. The first patient, with MDS, developed grade II aGVHD of the skin at Day 35 and relapsed at Day 105 with transition to hospice due to poor functional status. The second patient developed moderate cGVHD with skin, ocular, joint/fascia and genital tract involvement. The patient remains disease-free and is receiving investigational treatment for cGVHD. The third patient developed grade II aGVHD of the skin at Day 70 and extensive cGVHD with skin and oral involvement at 2-year. The patient remains disease and GVHD-free.

Note, CMV reactivation occurred in 18% of patients, and BK viruria occurred in 5% of patients. A list of non-hematological and non-infectious grades 3–4 toxicities up to Day 100 post-transplant is provided in Table S3.

Our findings indicate that the use of a modified strategy of one dose PTCy in myeloablative, unrelated, PBSC transplant results in a low incidence of severe aGVHD without compromising relapse risk. However, this regimen does not appear to mitigate the risk of cGVHD compared to standard pharmacologic prophylaxis, including two doses of PTCy.

The low incidence of severe aGVHD observed in our study is comparable to previous trials utilizing PTCy in various settings. The incidence of GVHD is reportedly higher with myeloablative allo-HSCT. Kanakry et al., reported a 60% CI of grades II-IV and 19% CI of grades III-IV aGVHD with two doses of PTCy, as a single agent, in myeloablative, matched-unrelated bone marrow transplant.² Luznik et al., using two doses of single agent PTCy in myeloablative matched-related and unrelated bone marrow transplant, reported an incidence of 43% and 10% for grades II-IV and III-IV aGVHD, respectively.³ Moiseev et al., reported a CI of 19% and 4% for grades II-IV and III-IV, respectively, with two doses of PTCy, tacrolimus, and MMF, in unrelated PBSC transplant. About 65% of the patients in that study received reduced-intensity conditioning, which was independently associated with a lower incidence of aGVHD.⁴ In comparison to the current literature, we show that one dose of PTCy was effective in controlling aGVHD in myeloablative, unrelated, PBSC transplants.

The rates of moderate and severe cGVHD in our study do not seem to be different than those observed with conventional pharmacologic GVHD prophylaxis with myeloablative PBSC transplants. On the other hand, results of studies with two doses of PTCy, in different platforms, have shown excellent control of cGVHD.^2-5 The discrepancy among studies is likely not only related to the number of PTCy doses, but also to the type of donors (related vs. unrelated), HLA-matching status, conditioning regimen intensity, and graft source (bone marrow vs. PBSC).

The 2-year relapse risk was 21% in our cohort, that included about one-third of patients with persistent disease at the time of transplant, and a large proportion with intermediate or high-risk disease. The reported incidence of relapse with PTCy has ranged from 19% to 46% using different conditioning intensity and donor sources.^2-5 The balance between impairing alloreactive T-cells and preserving Tregs with PTCy may determine the risk of both GVHD and relapse in this setting. However, several studies have now shown that there does not appear to be an increased risk of relapse with PTCy.⁵

Our reported NRM rate was comparable to prior reports of myeloablative PBSC transplant.⁴ Mortality before Day 100 in our study was primarily seen in patients >60 years old and attributed to sepsis rather than organ dysfunction. Beyond Day 100, cGVHD was the main cause of death. A recent CIBMTR analysis reported a 1-year NRM and cGVHD rate of 15% and 25%, respectively, in myeloablative, matched-unrelated donor transplant using two doses of PTCy.⁹ Extrapolating from these results, it is possible that two doses of PTCy would have resulted in lower NRM rates due to better cGVHD control compared to one dose, especially in a myeloablative, matched-unrelated, PBSC transplant.

We acknowledge the limitations of our study, being a single center with a relatively small sample size. Although we find one dose of PTCy to be effective in terms of aGVHD control and low incidence of relapse, we do not find a significant advantage with this strategy to improve the overall outcomes of myeloablative PBSC transplant and perhaps two doses are needed for better outcomes. To the best of our knowledge, our study is the first report to describe outcomes of myeloablative, matched-unrelated, PBSC transplant with one dose of PTCy. It serves as a platform to further investigate PTCy in larger such trials.

移植物抗宿主病 (GVHD) 是人类白细胞抗原 (HLA) 匹配的异基因造血干细胞移植 (allo-HSCT) 的主要并发症。药物 GVHD 预防通常包括钙调神经磷酸酶抑制剂、吗替麦考酚酯 (MMF)、甲氨蝶呤和西罗莫司。其他 GVHD 预防方法包括 T 细胞耗竭，然而，这些方法可能会增加感染和复发的风险，并可能导致免疫重建延迟的不同结果。^1-3

两种剂量的移植后环磷酰胺 (PTCy) 已显示出在半相合 (haplo)-HSCT 中作为 GVHD 预防的功效。从其在 haplo-HSCT 中的成功推断，PTCy 还证明了其在清髓性 HLA 匹配相关和无关骨髓移植中作为单一药物的有效性^，2、3以及与其他免疫抑制剂联合用于降低强度的外周血干细胞(PBSC) 移植。^{4, 5}此外，回顾性地发现 PTCy 与其他药物的组合在清髓性 PBSC 移植中是有效的。¹然而，在强度降低的情况下使用两剂 PTCy 进行的初步研究引起了对复发风险增加的担忧。⁶此外，PTCy 可能会增加感染并发症的风险，尤其是巨细胞病毒 (CMV) 感染。⁷为了平衡 GVHD 和移植物抗肿瘤效应，已经测试了一剂 PTCy。在非清髓性半相合骨髓移植中，与两剂相比，一剂有更高的广泛慢性 GVHD (cGVHD) 的趋势。感染发生率无显着差异。⁸一剂 PTCy 是否足以在清髓性、不相关的 PBSC 移植中预防 GVHD，目前尚不清楚。

我们进行了一项单臂 II 期临床试验，该试验使用一剂 PTCy、他克莫司和 MMF，用于在清髓性、不相关的 PBSC 异基因造血干细胞移植后预防 GVHD。该试验已在clinicaltrials.gov (NCT02065154) 上注册。该研究的主要终点是移植后第 100 天 II-IV 级急性 GVHD (aGVHD) 的累积发生率 (CI)。次要终点包括 cGVHD 的累积发生率、总生存期 (OS)、无进展生存期 (PFS)、复发和非复发相关死亡率 (NRM)。有关资格、准备方案、GVHD 预防、诊断和治疗、支持性护理和统计分析的详细信息在附录 S1 中提供。

2013 年 9 月至 2018 年 6 月期间招募了 39 名患者。 人口统计学和临床​​特征详见表 S1。36 位供者为 8/8 HLA 匹配无关，3 位为 7/8 匹配（A 错配 = 1，B 错配 = 1，C 错配 = 1）无关。该试验后来被修正，排除了 7/8 匹配的患者，这些患者的发展数据显示通过使用传统的两剂 PTCy 可以更有效地控制 GVHD。

在植入前的前 30 天内有 1 人死亡。其余 38 (97%) 名患者成功移植。ANC 和血小板植入的中位时间分别为 12 天（范围 9-14 天）和 33 天（范围 28-54 天）。图 S1A、S1B 分别提供了供体嵌合和免疫重建的 1 年数据。

移植后第 100 天 II-IV 和 III-IV 级 aGVHD 的 CI 为 30%（95% CI 15.2%–40.1%）（图 1A）和 5%（95% CI 2.1%–8.2%）（图1B），分别（表S2）。有两名患者为 IV 级 aGVHD（下消化道受累），没有一名患者为 III 级。轻度、中度和重度 cGVHD 的 2 年 CI 为 15% (95% CI 7.1%–23.3%)、15% (95% CI 8.4%–23.3%) 和 18% (95% CI 12.6%–27.4) ％）， 分别。对于 cGVHD 患者，皮肤是最常见的受累器官（n  = 9）。6 名患者有肝脏 GVHD（Bu/Flu = 6），而 4 名患者有肺部受累（Bu/Flu =3，Cy/TBI =1）。

急性和慢性 GVHD 的发生与诊断年龄（≥60 岁 vs. <60 岁和≥55 岁 vs. <55 岁）、性别、预处理方案（Bu/Flu vs. 其他、TBI）无显着差异与非 TBI）、疾病（AML 与其他）、DRI（高与其他）和 HLA 分型（8/8 与 7/8）。

复发和 NRM 的 2 年 CI 分别为 21%（95% CI 7.6%–33.4%）和 34%（95% CI 18.6%–49.6%）（图 S2）。在移植后第 100 天之前有 4 例非复发相关死亡。其中三名患者年龄> 60岁。一名患者在中性粒细胞移植前死亡。其余三名患者的死因是败血症（粪肠球菌 =1，表皮葡萄球菌 =1，生物体不明=1），导致多器官衰竭。对于移植后第 100 天后死亡的患者，最常见的死亡原因是 cGVHD（13%），其次是无潜在 GVHD 的感染（5%）。在死于 cGVHD 的患者中，主要的次要死亡原因包括由潜在肺 GVHD 和感染引起的呼吸衰竭。

PFS 和 OS 的 2 年估计值分别为 45% (95% CI 29.5%–61.3%) 和 51% (95% CI 34.7%–66.5%)。中位 OS 为 36 个月（95% CI 0-71.8 个月），中位随访时间为 50 个月（范围 6-71 个月）（图 S3）。

三名患者有 7/8 匹配的无关供体。第一位患有 MDS 的患者在第 35 天出现了 II 级皮肤 aGVHD，并在第 105 天复发，由于功能状态不佳而过渡到临终关怀。第二名患者发展为中度 cGVHD，皮肤、眼部、关节/筋膜和生殖道受累。该患者仍然无病，并正在接受 cGVHD 的研究性治疗。第三位患者在第 70 天出现皮肤 II 级 aGVHD，并在 2 年时出现皮肤和口腔受累的广泛 cGVHD。患者保持疾病和无 GVHD。

请注意，18% 的患者发生 CMV 再激活，5% 的患者发生 BK 病毒尿。表 S3 中提供了直至移植后第 100 天的非血液学和非传染性 3-4 级毒性列表。

我们的研究结果表明，在清髓性、无关的 PBSC 移植中使用一剂 PTCy 的改良策略可在不影响复发风险的情况下降低严重 aGVHD 的发生率。然而，与标准药物预防措施（包括两剂 PTCy）相比，该方案似乎并未降低 cGVHD 的风险。

在我们的研究中观察到的严重 aGVHD 的低发生率与之前在各种环境中使用 PTCy 的试验相当。据报道，清髓异基因造血干细胞移植的 GVHD 发生率更高。Kanakry 等人报道，在清髓性、不相关的骨髓移植中，使用两剂 PTCy 作为单一药物，II-IV 级 aGVHD 的 CI 为 60%，III-IV 级 aGVHD 的 CI 为 19%。² Luznik 等人在清髓性匹配相关和无关骨髓移植中使用两种剂量的单药 PTCy，报告称 II-IV 和 III-IV 级 aGVHD 的发生率分别为 43% 和 10%。³Moiseev 等人报道，在无关的 PBSC 移植中，使用两种剂量的 PTCy、他克莫司和 MMF，II-IV 和 III-IV 级的 CI 分别为 19% 和 4%。该研究中约 65% 的患者接受了强度降低的调理，这与 aGVHD 的较低发生率独立相关。⁴与当前文献相比，我们表明一剂 PTCy 可有效控制清髓性、无关的 PBSC 移植中的 aGVHD。

在我们的研究中，中度和重度 cGVHD 的发生率似乎与通过清髓性 PBSC 移植进行常规药物 GVHD 预防所观察到的发生率没有什么不同。另一方面，在不同平台上使用两种剂量的 PTCy 的研究结果表明，对 cGVHD 有极好的控制。^2-5研究之间的差异可能不仅与 PTCy 剂量的数量有关，还与供体类型（相关与无关）、HLA 匹配状态、预处理方案强度和移植物来源（骨髓与非相关）有关。 PBSC)。

在我们的队列中，2 年复发风险为 21%，其中包括大约三分之一在移植时患有持续疾病的患者，以及很大一部分患有中度或高风险疾病的患者。据报道，使用不同的调理强度和供体来源，PTCy 的复发率在 19% 到 46% 之间。^2-5损害同种异体反应性 T 细胞和用 PTCy 保留 Treg 之间的平衡可能决定在这种情况下 GVHD 和复发的风险。然而，现在的几项研究表明，PTCy 似乎不会增加复发的风险。⁵

我们报告的 NRM 率与先前关于清髓性 PBSC 移植的报告相当。⁴在我们的研究中，第 100 天之前的死亡率主要出现在 60 岁以上的患者中，并且归因于败血症而不是器官功能障碍。在第 100 天之后，cGVHD 是主要的死亡原因。最近的一项 CIBMTR 分析报告称，在使用两种剂量的 PTCy 进行清髓性匹配无关供体移植中，1 年 NRM 和 cGVHD 率分别为 15% 和 25%。⁹从这些结果推断，与一剂相比，两剂 PTCy 可能会导致较低的 NRM 率，因为与一剂相比，cGVHD 控制更好，尤其是在清髓性、匹配无关的 PBSC 移植中。

我们承认我们的研究存在局限性，因为它是一个样本量相对较小的单一中心。尽管我们发现一剂 PTCy 在控制 aGVHD 和低复发率方面是有效的，但我们没有发现这种策略在改善清髓性 PBSC 移植的总体结果方面具有显着优势，也许需要两剂才能获得更好的结果。据我们所知，我们的研究是第一份描述清髓性、不相关的 PBSC 移植与一剂 PTCy 的结果的报告。它作为一个平台，可以在更大的此类试验中进一步研究 PTCy。