Pooled allogeneic faecal microbiota MaaT013 for steroidresistant gastrointestinal acute graft-versus-host disease: a single-arm, multicentre phase 2 trial 治疗类固醇耐药的胃肠道急性移植物抗宿主病的集合异体粪便微生物群MaaT013：单臂、多中心2期试验
Florent Malard, ^(a,**){ }^{a, *} Michael Loschi, ^(b){ }^{b} Anne Huynh, ^(c){ }^{c} Thomas Cluzeau, ^(b){ }^{b} Sarah Guenounou, ^(c){ }^{c} Faezeh Legrand, ^(d){ }^{d} Leonardo Magro, ^(e){ }^{e} Corentin Orvain, ^(f){ }^{f} Amandine Charbonnier, ^(9){ }^{9} Marta Panz-Klapuch, ^(h){ }^{h} Deborah Desmier,' Jean-Baptiste Mear,, Jérôme Cornillon, ^(k){ }^{k} Christine Robin,, Etienne Daguindau, ^(m){ }^{m} Karin Bilger, ^(n){ }^{n} Maria J. G. T. Vehreschild, ^(@){ }^{\circ} Patrice Chevallier, ^(p){ }^{p} Hélène Labussière-Wallet, ^(q){ }^{q} Clémence Mediavilla, ^(r){ }^{r} Marie-Anne Couturier, ^(5){ }^{5} Claude-Eric Bulabois, ^(t){ }^{t} Vincent Camus, ^("U "){ }^{\text {U }} Sylvain Chantepie, ^(v){ }^{v} Patrice Ceballos, ^(,){ }^{,}Béatrice Gaugler, ^(a){ }^{a} Ernst Holler, ^(x){ }^{x} Joël Doré, Emmanuel Prestat, ^(2){ }^{2} Cyrielle Gasc, ^(z){ }^{z} Emilie Plantamura, ^(z){ }^{z} and Mohamad Mohty ^(a){ }^{a} Florent Malard, ^(a,**){ }^{a, *} Michael Loschi, ^(b){ }^{b} Anne Huynh, ^(c){ }^{c} Thomas Cluzeau, ^(b){ }^{b} Sarah Guenounou, ^(c){ }^{c} Faezeh Legrand, ^(d){ }^{d} Leonardo Magro, ^(e){ }^{e} Corentin Orvain、 ^(f){ }^{f} Amandine Charbonnier、 ^(9){ }^{9} Marta Panz-Klapuch、 ^(h){ }^{h} Deborah Desmier、' Jean-Baptiste Mear、Jérôme Cornillon、 ^(k){ }^{k} Christine Robin、Etienne Daguindau、 ^(m){ }^{m} Karin Bilger、 ^(n){ }^{n} Maria J．G. T.Vehreschild, ^(@){ }^{\circ} Patrice Chevallier, ^(p){ }^{p} Hélène Labussière-Wallet, ^(q){ }^{q} Clémence Mediavilla, ^(r){ }^{r} Marie-Anne Couturier, ^(5){ }^{5} Claude-Eric Bulabois, ^(t){ }^{t} Vincent Camus, ^("U "){ }^{\text {U }} Sylvain Chantepie、 ^(v){ }^{v} Patrice Ceballos、 ^(,){ }^{,} Béatrice Gaugler、 ^(a){ }^{a} Ernst Holler、 ^(x){ }^{x} Joël Doré、Emmanuel Prestat、 ^(2){ }^{2} Cyrielle Gasc、 ^(z){ }^{z} Emilie Plantamura 和 Mohamad Mohty。^(a){ }^{a} Sorbonne Université, AP-HP, Centre de Recherche Saint-Antoine INSERM UMRs938, Service D'Hématologie Clinique et de Thérapie Cellulaire, Hôpital Saint Antoine, AP-HP, Paris, France ^(a){ }^{a} 法国巴黎索邦大学，AP-HP，圣安托万研究中心 INSERM UMRs938，圣安托万医院，AP-HP，细胞病理学临床与治疗服务部。
^("c"){ }^{\text {c}} Service Hématologie, CHU/IUCT-Oncopole, Toulouse Cédex 31059, France ^("c"){ }^{\text {c}} 法国图卢兹Cédex 31059，CHU/IUCT-Oncopole肿瘤医院肿瘤科dHaematology Department, Institut Paoli Calmettes, Marseille, France d 法国马赛保利卡尔梅特研究所血液学部é Unité d'Allogreffe, Maladies du sang, CHRU, Lille 59000, France 法国里尔，59000，CHRU，血液病，同种异体移植组^(f){ }^{f} Department of Haematology, CHU d'Angers, France ^(f){ }^{f} 法国昂热大学血液学系^("g"){ }^{\text {g}} Service d'Hématologie Clinique, CHU Amiens-Picardie, Amiens, France ^("g"){ }^{\text {g}} 亚眠-皮卡第慢性病医院临床病理科，法国亚眠^(h){ }^{h} Department of Haematology and Bone Marrow Transplantation, School of Medicine in Katowice, Medical University of Silesia, Dąbrowski Street, 25, Katowice 40-032, Poland ^(h){ }^{h} 西里西亚医科大学卡托维兹医学院血液学和骨髓移植系，地址：Dąbrowski Street, 25, Katowice 40-032, Poland^(i)CHU{ }^{i} \mathrm{CHU} de Poitiers, Service d'Hématologie et de Thérapie Cellulaire, Poitiers, France ^(i)CHU{ }^{i} \mathrm{CHU} 法国普瓦捷，血液学和细胞治疗部，法国普瓦捷^(j){ }^{j} Clinical Haematology, University Hospital of Rennes, Rennes, France ^(j){ }^{j} 雷恩大学医院临床血液学，法国雷恩képartement d'Hématologie Clinique et de Thérapie Cellulaire, CHU de St-Etienne, Saint-Etienne, France'Hôpital Henri Mondor, Service d'Hématologie Clinique et de Thérapie Cellulaire, Créteil, France 法国克雷泰尔亨利-蒙多尔医院，细胞病理学临床与治疗服务部mépartement d'Hématologie, CHU de Besançon, Besançon, FrancenHôpital de Hautepierre, Pôle Oncologie-Hématologie, Strasbourg, France 法国斯特拉斯堡 Hautepierre 医院肿瘤-血细胞学科^(@){ }^{\circ} Department I of Internal Medicine, University Hospital of Cologne, Cologne, Germany ^(@){ }^{\circ} 科隆大学医院内科 I 部，德国科隆^(p){ }^{p} Clinical Haematology, Nantes University Hospital, Nantes, France ^(p){ }^{p} 临床血液学，南特大学医院，法国南特^(q){ }^{q} Hôpital Lyon Sud, Hospices Civils de Lyon, Pierre Bénite, France ^(q){ }^{q} 里昂南方医院，里昂平民医院，皮埃尔-贝尼特，法国'Haematology Department, CHU de Bordeaux, Bordeaux, France 法国波尔多 CHU de Bordeaux 血液科^(5){ }^{5} Department of Haematology, CHRU Brest, Brest, France ^(5){ }^{5} 法国布雷斯特CHRU血液学系^(t)CHU{ }^{\mathrm{t}} \mathrm{CHU} Grenoble Alpes - Université Grenoble Alpes, Haematology, Grenoble, France ^(t)CHU{ }^{\mathrm{t}} \mathrm{CHU} 格勒诺布尔阿尔卑斯 - 格勒诺布尔阿尔卑斯大学，血液学，法国格勒诺布尔"Department of Haematology and INSERM U1245, Centre Henri Becquerel, Roven, France "法国罗文亨利-贝克勒尔中心血液学系和 INSERM U1245^(v){ }^{v} Institut d'Hématologie de Basse Normandie, CHU Caen Normandie, Caen, France ^(v){ }^{v} 法国卡昂，卡昂诺曼底 CHU 下诺曼底血液学研究所^(w)CHU{ }^{w} \mathrm{CHU} de Montpellier, Hôpital Saint Eloi, Montpellier, France ^(w)CHU{ }^{w} \mathrm{CHU} 蒙彼利埃，圣埃洛伊医院，蒙彼利埃，法国^(x){ }^{x} Department of Internal Medicine III, University Medical Centre, Regensburg, Germany ^(x){ }^{x} 德国雷根斯堡大学医学中心内科三部^(y){ }^{y} INRAE, MGP, Université Paris-Saclay, Jouy-en-Josas 78350, France ^(y){ }^{y} INRAE, MGP, Université Paris-Saclay, Jouy-en-Josas 78350, France^(z){ }^{\mathrm{z}} MaaT Pharma, Lyon, France ^(z){ }^{\mathrm{z}} 法国里昂 MaaT 制药公司
Summary 摘要
Background Failure of gastrointestinal acute graft-versus-host disease (GI-aGvHD) to respond to steroid therapy is associated with limited further therapeutic options. We aimed to assess the safety and efficacy of the first-in-human use of the pooled allogeneic faecal microbiota, MaaT013, for the treatment of steroid-refractory GI-aGvHD. 背景胃肠道急性移植物抗宿主病（GI-aGvHD）对类固醇治疗无效，进一步的治疗选择有限。我们旨在评估首次在人体中使用集合异体粪便微生物群 MaaT013 治疗类固醇难治性 GI-aGvHD 的安全性和有效性。
Methods This prospective, international, single-arm, phase 2a study reports clinical outcomes from a 24-patient cohort with grade III-IV, steroid refractory GI-aGvHD treated with the pooled allogeneic faecal microbiota MaaT013. MaaT013 involved pooling faecal matter from 3 to 8 screened donors then transplanting the pooled batches into patients to treat GI-aGVHD. The 24 patients were treated in the HERACLES study (Aug 2018 to Nov 2020) at 26 sites in Europe and an additional 52 patients were treated in a compassionate use/expanded access program (EAP) in France (July 2018 to April 2021). The primary endpoint was GI response at day 28 , defined as the proportion of patients with GI-aGvHD who had a complete response (CR) or very good partial response (VGPR). GvHD grading and staging were assessed according to 方法 这项前瞻性、国际性、单臂 2a 期研究报告了 24 例 III-IV 级、类固醇难治性 GI-aGvHD 患者使用集合异体粪便微生物群 MaaT013 治疗的临床结果。MaaT013 是将 3 到 8 名经过筛选的供体的粪便汇集在一起，然后将汇集的粪便移植到患者体内以治疗 GI-aGVHD。24名患者在欧洲26个研究机构的HERACLES研究中接受治疗（2018年8月至2020年11月），另有52名患者在法国的一项同情使用/扩大准入计划（EAP）中接受治疗（2018年7月至2021年4月）。主要终点是第28天的消化道反应，定义为完全反应（CR）或非常好的部分反应（VGPR）的消化道GvHD患者比例。GvHD分级和分期根据以下标准进行评估
eClinicalMedicine 2023;62: 102111
the revised Glucksberg criteria. Adverse events and severe adverse events were monitored for 6 months and 12 months, respectively. The HERACLES study was registered with ClinicalTrials.gov (NCT03359980). 修订后的 Glucksberg 标准。不良事件和严重不良事件的监测时间分别为6个月和12个月。HERACLES研究已在ClinicalTrials.gov（NCT03359980）上注册。
Findings Compared with single donors, MaaT013 is characterised by higher microbial richness and reduced variability across batches. At day 28 (D28), the GI-overall response rate (ORR) was 38%38 \% in the prospective population, including 5 complete responses (CR), 2 very good partial responses (VGPR) and 2 partial responses (PR). In the EAP, the GI-ORR was 58%58 \% ( 17 CR, 9 VGPR and 4 PR). The 12-month overall survival (OS) was 25%25 \% in the prospective study and 38%38 \% in the EAP. Regarding safety, five infectious complications, including 3 sepsis, could not be excluded from being related to the study procedure in HERACLES. Shotgun sequencing analyses of the identified strains suggest that none were found in MaaT013. In the EAP, 18 pharmacovigilance cases were reported among 52 treated patients, including 11 bacteraemia/sepsis. In HERACLES, we observed in stools from responding patients at D28 a higher microbiota richness and increased levels of beneficial bacteria, in particular butyrate producers, along with increased levels of short-chain fatty acid and bile acids. In contrast, stools from non-responding (NR) patients displayed increased levels of pathogenic pro-inflammatory bacteria along with increased systemic inflammatory parameters. 研究结果 与单一供体相比，MaaT013 的特点是微生物丰富度更高，不同批次之间的差异更小。在第28天（D28），前瞻性人群的胃肠道总体反应率（ORR）为 38%38 \% ，包括5例完全反应（CR）、2例非常好的部分反应（VGPR）和2例部分反应（PR）。在EAP中，消化道ORR为 58%58 \% （17例CR、9例VGPR和4例PR）。在前瞻性研究中，12个月的总生存期（OS）为 25%25 \% ，在EAP中为 38%38 \% 。在安全性方面，在HERACLES研究中，5例感染性并发症（包括3例败血症）不能排除与研究过程有关。对确定的菌株进行的射枪测序分析表明，MaaT013 中未发现任何菌株。在 EAP 中，52 名接受治疗的患者中有 18 例药物警戒病例，其中包括 11 例菌血症/败血症。在 HERACLES 中，我们观察到，在第 28 天时，应答患者的粪便中微生物群丰富度提高，有益菌水平增加，尤其是丁酸菌，同时短链脂肪酸和胆汁酸水平增加。与此相反，无反应（NR）患者的粪便中致病性促炎症细菌的含量增加，同时全身炎症参数升高。
Interpretation Overall, MaaT013 was safe in this population of highly immunocompromised patients and was associated with responses in some patients with GI-aGvHD and deserves further investigation. 释义 总体而言，MaaT013 在这一免疫功能高度低下的患者群体中是安全的，并且与一些消化道- aGvHD 患者的反应有关，值得进一步研究。
Research in context 研究背景
Evidence before this study 本研究之前的证据
To assess previous research pertaining to faecal microbiota transplantation for gastrointestinal acute graft-versus-host disease (GI-aGvHD) following allogeneic haematopoietic cell transplantation (allo-HCT), we conducted a non-systematic PubMed search using a combination of terms including, but not limited to, 'Faecal microbiota transplantation, ‘graft-versus-host disease’, and ‘allogeneic haematopoietic cell transplantation’. The search included clinical trials and clinical observations (with no start date up to Feb 28, 2023). While faecal microbiota transplantation is a well-established treatment for recurrent Clostridioides difficile colitis, only case reports and retrospective analyses are available regarding is use for treatment of Gl-aGvHD\mathrm{Gl}-\mathrm{aGvHD}. These reports suggest that faecal microbiota transplantation is safe in immunocompromised patients and may be effective. Nevertheless, faecal microbiota transplantation for GI-aGvHD was never investigated in a prospective study with clearly define inclusion and exclusion criteria. Thus, there are insufficient data to inform on the safety and efficacy of faecal microbiota transplantation for steroid refractory Gl-aGvHD\mathrm{Gl}-\mathrm{aGvHD}. 为了评估有关粪便微生物群移植治疗异基因造血细胞移植（allo-HCT）后胃肠道急性移植物抗宿主疾病（GI-aGvHD）的既往研究，我们使用包括但不限于 "粪便微生物群移植"、"移植物抗宿主疾病 "和 "异基因造血细胞移植 "的术语组合进行了非系统性的 PubMed 搜索。搜索范围包括临床试验和临床观察（无开始日期，截至 2023 年 2 月 28 日）。虽然粪便微生物群移植是治疗复发性艰难梭菌性结肠炎的一种行之有效的方法，但只有病例报告和回顾性分析可用于治疗 Gl-aGvHD\mathrm{Gl}-\mathrm{aGvHD} 。这些报告表明，粪便微生物群移植对免疫力低下的患者是安全的，而且可能有效。然而，粪便微生物群移植治疗消化道-aGvHD从未在明确界定纳入和排除标准的前瞻性研究中进行过调查。因此，目前还没有足够的数据来说明粪便微生物群移植治疗类固醇难治性 Gl-aGvHD\mathrm{Gl}-\mathrm{aGvHD} 的安全性和有效性。
Added value of this study 本研究的附加值
To address this knowledge gap, we conducted a multi-centre, prospective, phase II study evaluating the pooled allogeneic faecal microbiota MaaT013, for steroid refractory GI-aGvHD. In a cohort of 24 steroid refractory GI -aGvHD, at day 28 , the Gl-overall response rate was 38%38 \% including 5 complete 为了填补这一知识空白，我们开展了一项多中心、前瞻性的II期研究，评估了集合异体粪便微生物群MaaT013治疗类固醇难治性消化道-aGvHD的效果。在24例类固醇难治性消化道-aGvHD患者中，第28天的Gl-总反应率为 38%38 \% ，其中包括5例完全缓解的消化道-aGvHD患者。
responses, 2 very good partial responses and 2 partial responses. We observed in stools from responding patients at day 28 a higher microbiota richness and increased levels of beneficial bacteria, in particular butyrate producers, while stool from non-responding patients displayed increased levels of pathogenic pro-inflammatory bacteria along with increased systemic inflammatory parameters. Finally, 52 additional patients with steroid refractory or steroid dependant GI aGvHD were also treated within a compassionate use/ expanded access program. In those patients the day 28Gl-28 \mathrm{Gl}- overall response rate was 58%. Regarding safety, five infectious complications, including 3 sepsis, could not be excluded from being related to the study procedure in HERACLES. Shotgun sequencing analyses of the identified strains suggest that none were found in MaaT013. In the EAP, 18 pharmacovigilance cases were reported among 52 treated patients, including 11 bacteraemia/sepsis. 我们观察到，在第 28 天，有反应的患者粪便中微生物群丰富度提高，有益菌水平增加，尤其是丁酸菌。我们观察到，在第 28 天时，应答患者粪便中的微生物群丰富度提高，有益菌（尤其是丁酸菌）的含量增加，而非应答患者粪便中的致病性促炎菌含量增加，全身炎症指标升高。最后，另有 52 名类固醇难治性或类固醇依赖性消化道 aGvHD 患者也在同情使用/扩大准入计划内接受了治疗。这些患者的日 28Gl-28 \mathrm{Gl}- 总反应率为58%。在安全性方面，有5例感染性并发症（包括3例败血症）不能排除与HERACLES研究过程有关。对确定的菌株进行的射枪测序分析表明，MaaT013 中未发现任何菌株。在 EAP 中，52 名接受治疗的患者中有 18 例药物警戒病例，包括 11 例菌血症/败血症。
Implications of all the available evidence 所有现有证据的影响
This is the first prospective study evaluating MaaT013 for treatment of steroid refractory Gl-aGvHD\mathrm{Gl}-\mathrm{aGvHD}. Along with previously reported case reports and retrospective studies, our findings show that the pooled allogeneic faecal microbiota MaaT013 is safe in the studied population of highly immunocompromised patients after allo-HCT. Some patients responded to faecal microbiota transplantation, indicating that it deserves further investigation. 这是第一项评估MaaT013治疗类固醇难治性 Gl-aGvHD\mathrm{Gl}-\mathrm{aGvHD} 的前瞻性研究。与之前报道的病例报告和回顾性研究一样，我们的研究结果表明，在所研究的异体肝移植后免疫力高度低下的患者群体中，汇集的异体粪便微生物菌群MaaT013是安全的。一些患者对粪便微生物群移植有反应，这表明它值得进一步研究。
Introduction 导言
Severe gastrointestinal acute graft-versus-host disease (GIaGvHD ) remains a major source of morbidity and mortality following allogeneic haematopoietic cell transplantation (allo-HCT). ^(1,2){ }^{1,2} Despite initial responses to first line treatment with corticosteroids (CS), ^(3,4){ }^{3,4} fewer than half of patients achieve a durable complete response (CR), and those who do not respond or progress after an initial response have high mortality. ^(5-7){ }^{5-7} For these reasons, there is great interest in identifying effective second-line therapies for steroidrefractory (SR)-GI-aGvHD to improve outcomes. 严重的胃肠道急性移植物抗宿主疾病（GIaGvHD）仍然是异基因造血细胞移植（allo-HCT）后发病率和死亡率的主要原因。 ^(1,2){ }^{1,2} 尽管皮质类固醇（CS）一线治疗取得了初步反应，但 ^(3,4){ }^{3,4} 只有不到一半的患者获得了持久的完全反应（CR），而那些在初步反应后没有反应或病情进展的患者死亡率很高。 ^(5-7){ }^{5-7} 由于这些原因，人们对确定治疗类固醇难治性（SR）-GI-aGvHD的有效二线疗法以改善预后非常感兴趣。
It is now well established that the intestinal bacterial microbiome undergoes profound changes during the allo-HCT procedure leading to alteration of patients’ outcomes. ^(8,9){ }^{8,9} In particular, loss of bacterial diversity is associated with GvHD-related outcomes after alloHCT. ^(10,11){ }^{10,11} Domination of Enterococcus after allo-HCT is associated with an increased GvHD-related mortality. ^(12){ }^{12} In contrast, Clostridiales, such as Blautia, have antiinflammatory homeostatic roles, including upregulation of regulatory T (Treg) cells through the production of the short-chain fatty acid (SCFA) butyrate, and are associated with a decreased GvHD-related mortality. ^(13,14){ }^{13,14} 肠道细菌微生物组在同种异体血细胞移植过程中发生深刻变化，导致患者预后改变，这一点现已得到公认。 ^(8,9){ }^{8,9} 特别是，细菌多样性的丧失与同种异体血细胞移植后的GvHD相关结果有关。 ^(10,11){ }^{10,11} allo-HCT术后肠球菌占优势与GvHD相关死亡率增加有关。 ^(12){ }^{12} 相反，梭状芽孢杆菌（如布劳氏菌）具有抗炎平衡作用，包括通过产生短链脂肪酸（SCFA）丁酸来上调调节性T（Treg）细胞，并与GvHD相关死亡率的降低有关。 ^(13,14){ }^{13,14}
Faecal microbiota transplantation (FMT), a wellestablished treatment for recurrent CC. difficile colitis, ^(15){ }^{15} has therefore been investigated in patients with GI-aGvHD. Several pilot studies have shown promising results of FMT in improving diarrhea and symptoms in patients with GI-GvHD and the procedure appears to be safe, despite the immunocompromised status of allo-HCT patients. ^(16-18){ }^{16-18} 粪便微生物群移植（FMT）是治疗复发性艰难梭菌性结肠炎 CC 的一种行之有效的方法， ^(15){ }^{15} 因此也被用于研究消化道-胃食管返流病患者。几项试点研究显示，FMT 在改善胃肠道 GvHD 患者的腹泻和症状方面取得了可喜的成果，尽管异体肝移植患者的免疫功能低下，但这种治疗方法似乎是安全的。 ^(16-18){ }^{16-18}
Based on these data, we designed a phase IIa clinical trial, to evaluate first-in-human use of the pooled allogeneic faecal microbiota MaaT013 for the treatment of SR-GI-aGvHD in 24 patients. In addition, 52 patients with SR or steroid-dependent (SD) GI-aGvHD were treated in a multicentre compassionate use/expanded access program (EAP) in France. 基于这些数据，我们设计了一项 IIa 期临床试验，评估首次在 24 名患者中使用集合异体粪便微生物群 MaaT013 治疗 SR-GI-aGvHD 的效果。此外，52 名 SR 或类固醇依赖型 (SD) GI-aGvHD 患者在法国的一项多中心同情使用/扩大使用计划 (EAP) 中接受了治疗。
Methods 方法
HERACLES clinical trial design HERACLES 临床试验设计
The HERACLES study (clinical trial NCT03359980), is a single-arm, open-label, non-randomised, multicentre (26 sites), phase IIa trial, to assess the safety and efficacy of MaaT013 for treating patients with grade III-IV, gut predominant, SR-aGvHD. Full eligibility criteria are provided in the Supplementary file and in Supplementary Table S1. The trial was conducted in accordance with the guidelines for Good Clinical Practice of the International Council for Harmonisation, with applicable local regulations, and according to the principles of the Declaration of Helsinki. The protocol (online only) was approved at each participating center by the relevant independent ethics committees and is available in the Supplementary Materials. HERACLES研究（临床试验NCT03359980）是一项单臂、开放标签、非随机、多中心（26个研究点）IIa期试验，旨在评估MaaT013治疗III-IV级肠道占位SR-aGvHD患者的安全性和有效性。全部资格标准见补充文件和补充表 S1。试验按照国际协调委员会的《良好临床实践指南》、当地适用法规以及《赫尔辛基宣言》的原则进行。每个参与中心的试验方案（仅在线版）均已获得相关独立伦理委员会的批准，可在补充材料中查阅。
HERACLES eligibility criteria HERACLES 的资格标准
Eligible patients were adult recipients of allo-HCT (any donor and any stem cell source, any GvHD 符合条件的患者均为接受过异体供体间充质干细胞移植的成年受者（任何供体和任何干细胞来源，任何 GvHD
prophylaxis or conditioning regimen), who developed a first episode of grade III-IV aGvHD with gut predominance (if any other organs were involved), resistant to first-line systemic therapy with CS. For patients with other organ involvement, steroidresistance of lower GI involvement was mandatory. Steroid-resistance was defined as progression after 3 days or a lack of improvement after 5 days of treatment with CS at 2mg//kg//d2 \mathrm{mg} / \mathrm{kg} / \mathrm{d} methylprednisolone equivalent dose. Patients treated with 1mg//kg//1 \mathrm{mg} / \mathrm{kg} / d of CS (because the physician deemed that they would not tolerate 2mg//kg//d2 \mathrm{mg} / \mathrm{kg} / \mathrm{d} ) and who corresponded to the definition of SR patients, could be included. Patients were excluded if they had grade IV hyperacute GvHD, aGvHD after donor lymphocyte infusion (DLI), or overlapping chronic GvHD; if they had a relapsed/persistent malignancy requiring rapid immune suppression withdrawal; if they had an active uncontrolled infection, current or past veno-occlusive disease or other uncontrolled complication (complete list of criteria in Supplementary Table S1). 预防或调理方案），首次出现以肠道为主的 III-IV 级 aGvHD（如果有其他器官受累），对 CS 一线全身治疗耐药。对于有其他器官受累的患者，必须是下消化道受累的类固醇耐药患者。类固醇耐药的定义是，使用 2mg//kg//d2 \mathrm{mg} / \mathrm{kg} / \mathrm{d} 甲基强的松龙等效剂量的CS治疗3天后病情进展或5天后无改善。接受 1mg//kg//1 \mathrm{mg} / \mathrm{kg} / d CS治疗的患者（因为医生认为他们无法耐受 2mg//kg//d2 \mathrm{mg} / \mathrm{kg} / \mathrm{d} ），如果符合SR患者的定义，也可纳入其中。如果患者患有IV级急性并发症、供体淋巴细胞输注（DLI）后并发症或重叠性慢性并发症；如果患者患有复发/顽固性恶性肿瘤，需要快速停用免疫抑制；如果患者患有未控制的活动性感染、目前或过去的静脉闭塞性疾病或其他未控制的并发症，则排除在外（标准的完整列表见补充表S1）。
HERACLES study treatment HERACLES 研究治疗
Each MaaT013 batch is manufactured in a French facility following current good manufacturing practice by pooling faecal material from 3 to 8 strictly vetted, healthy donors. The safety testing strategy comprises medical evaluation, regular testing of blood and faeces following current regulatory recommendations for safety testing ^(19){ }^{19} (Supplementary Table S2 and S3). MaaT013 manufacturing relies on the simultaneous availability of multiple healthy, qualified donors and stools from donors must be manufactured within 72 h after collection. Due to logistical constraints, MaaT013 manufacturing is performed periodically over a specified time window of several weeks, referred to as a “manufacturing campaign”. Each manufacturing campaign includes qualification of healthy donors, daily stool collection, batch manufacturing donor health close follow-up, and product quality control leading to batch release. Each stool from a participating donor is mixed independently from the others with a cryopreservative diluent at a ratio of 4 mL to 1 g of faeces, and filtered-this is referred to as a single donor suspension. Suspensions from 3 to 8 single donors are then extemporaneously pooled and mixed-each pool being referred to as one batch (number of batches produced per day being dependent on the amount of collected stools). The pooled suspension is then distributed into 150 mL freeze-resistant bags (one bag being equivalent to one dose of MaaT013) and stored at -80^(@)C-80^{\circ} \mathrm{C}. Pooling allows the standardisation of the product composition and intra-batch consistency regarding relative abundance of the main phyla including specific genera associated with clinical benefits, such as butyrate-producing bacterial genera. Complete information on MaaT013 manufacturing and characterisation is 每一批 MaaT013 都是在法国的一家工厂按照现行的良好生产规范，将 3 至 8 名经过严格审查的健康供体的粪便材料汇集在一起生产的。安全检测策略包括医学评估、血液和粪便的定期检测，并遵循当前监管机构对安全检测的建议 ^(19){ }^{19} （补充表 S2 和 S3）。MaaT013 的生产需要同时获得多名健康、合格的供体，供体的粪便必须在采集后 72 小时内生产。由于后勤方面的限制，MaaT013 的生产需要在几周的特定时间窗口内定期进行，称为 "生产活动"。每次生产活动包括对健康供体的资格审查、每日粪便采集、批量生产供体健康状况的密切跟踪，以及产品质量控制，直至批量发布。每个参与供体的粪便都与其他供体的粪便按 4 毫升兑 1 克粪便的比例单独与低温保存稀释液混合，然后过滤--这被称为单个供体悬浮液。然后将 3 至 8 个单一供体的悬浮液临时汇集并混合，每个汇集的悬浮液称为一个批次（每天生产的批次数量取决于收集的粪便量）。然后将汇集的悬浮液分装到 150 mL 的抗冻袋中（一袋相当于一剂 MaaT013），并储存在 -80^(@)C-80^{\circ} \mathrm{C} 中。汇集悬浮液可使产品成分标准化，并使批内主要菌属的相对丰度保持一致，包括与临床益处相关的特定菌属，如产生丁酸的细菌菌属。有关 MaaT013 生产和表征的完整信息，请参见
available in the Supplementary file and in Supplementary Fig. S1. 见补充文件和补充图 S1。
The first MaaT013 administration was performed within 5 days of patient inclusion. A total of 3 administrations ( 7+-27 \pm 2 days between each administration) were planned. The administration was performed by thawing the product and then administering via a rectal catheter. Faecal and blood samples were collected from D0 to D28: at D0, D9, D16, and D28. When collection coincided with the day of MaaT013 administration (D0, D9, and D16), blood and faeces were always collected before MaaT013 administration. Guidelines were provided for steroid tapering based on available recommendations (see Supplementary Material). ^(3){ }^{3} No additional systemic second-line drugs were allowed during MaaT013 treatment. 首次给药在患者入院后 5 天内进行。计划总共给药 3 次（每次给药间隔 7+-27 \pm 2 天）。给药方法是将产品解冻，然后通过直肠导管给药。从第 0 天到第 28 天，分别在第 0 天、第 9 天、第 16 天和第 28 天采集粪便和血液样本。如果采集时间与服用 MaaT013 的时间一致（D0、D9 和 D16），则在服用 MaaT013 之前采集血液和粪便样本。根据现有建议提供了类固醇减量指南（见补充材料）。 ^(3){ }^{3} 在MaaT013治疗期间，不允许额外使用全身性二线药物。
Endpoints and assessments 终点和评估
The primary endpoint was GI response at D28, which was defined as the proportion of patients with GIaGvHD who had a CR or very good partial response (VGPR). The endpoint was not achieved for patients who (1) did not achieve at least GI-VGPR at D28, (2) achieved at least GI-VGPR before D28 and did not have a durable CR/VGPR on D28, (3) received additional systemic GvHD therapy before D28, or (4) died before D28. CR was defined as complete resolution of GIaGvHD manifestations. VGPR was defined as improvement of at least 2 stages in the severity of GIaGvHD , or improvement of the GI staging from 2 to 1 , except improvement to stage 0 . GvHD grading and staging were performed by the investigators according to the revised Glucksberg criteria. ^(20){ }^{20} 主要终点是 D28 时的 GI 反应，定义为 GIaGvHD 患者中获得 CR 或非常好的部分反应 (VGPR) 的比例。以下患者未达到终点：(1) D28时至少未达到GI-VGPR；(2) D28前至少达到GI-VGPR，但D28时未达到持久CR/VGPR；(3) D28前接受了额外的全身性GvHD治疗；或(4) D28前死亡。CR 定义为 GIaGvHD 表现完全消失。VGPR 的定义是 GIaGvHD 的严重程度至少改善 2 期，或 GI 分期从 2 期改善到 1 期，但改善到 0 期除外。研究人员根据修订后的 Glucksberg 标准对 GvHD 进行分级和分期。 ^(20){ }^{20}
Secondary endpoints included the best response rates (CR, VGPR, and partial response [PR-defined as improvement of 1 stage in the severity of GI-aGvHD]) for GI and overall aGvHD , the duration of response, the safety, the feasibility and MaaT013 activity and impact on overall survival (OS), and evaluation of microbiota reconstitution post MaaT013 administration. Safety, adverse events (AEs), and severe AEs (SAEs) were monitored until 6 months for AEs and 12 months for SAEs according to the procedure detailed in the Supplementary file. 次要终点包括消化道和整体 aGvHD 的最佳反应率（CR、VGPR 和部分反应[PR 定义为消化道 aGvHD 的严重程度改善一个阶段]）、反应持续时间、安全性、MaaT013 活性的可行性和对总生存期 (OS) 的影响，以及评估 MaaT013 用药后微生物群重建情况。安全性、不良事件（AEs）和严重不良事件（SAEs）的监测时间分别为6个月和12个月，具体过程详见补充文件。
Statistical analysis 统计分析
As the study was a single-arm, open-label trial, no randomisation nor blinding was performed. As the objective of the study was to assess both crude rate of response and safety profile of MaaT013 in SR-GIaGvHD patients, no formal sample size calculation was made. Recruitment of 32 patients was initially planned. During the COVID-19 pandemic, the trial was put on-hold by the French Agency for the Safety of Health Products (ANSM) and recruitment stopped in March 2020 with 24 patients treated out of the planned 32 patients. 由于该研究是一项单臂、开放标签试验，因此没有进行随机分组或盲法。由于研究的目的是评估MaaT013在SR-GIaGvHD患者中的粗反应率和安全性，因此没有进行正式的样本量计算。最初计划招募 32 名患者。在COVID-19大流行期间，法国保健品安全局（ANSM）暂停了该试验，并于2020年3月停止了招募，在计划招募的32名患者中，有24名患者接受了治疗。
OS was calculated by the Kaplan-Meier method. OS was defined as the time from D0 to death, regardless of the cause. All p values are 2 -sided and any value of p < 0.05p<0.05 was considered statistically significant. For microbiome comparison between single healthy donors and MaaT013 batches, we used Fisher’s F test for equality of variance. The Bray-Curtis similarity index ( 1 minus the Bray-Curtis dissimilarity index) was calculated for donors (all campaigns) and MaaT013 batches per manufacturing campaign. We tested the homogeneity of variances and the equality of medians of the similarity measures between single healthy donors and MaaT013 batches for each of the five campaigns using Fischer and Wilcoxon tests, respectively (for both tests, we adjusted the p-values according to the False Discovery Rate (FDR) ^(21){ }^{21} for the five comparisons). To assess the difference in richness between single donors and pools, we used a t-test with Welch’s correction for heteroscedasticity. For microbiome and faecal metabolites comparison, to compare the responding ( R , patients in CR,VGPR\mathrm{CR}, \mathrm{VGPR} or PR at D 28 ) and non-responding ( NR , patients not in CR, VGPR or PR at D28) groups at D0, 9, 16 , and 28 , as well as to compare R groups between D0 and D9, D0 and D16, and D0 and D28, 7 unpaired twosided Wilcoxon tests were performed. For engraftment, the percentage of engrafted operational taxonomic units (OTU) were compared at D9, 16, and 28 between RR and NR groups for OTUs identified before treatment in the product only, in the patient only, or for both product and patient. The p -value was considered significant after FDR controlled multiple tests ( 3 tests for engraftment metrics, 7 tests for microbiome and faecal metabolite comparison) adjustment when it is lower than 0.05 (the p -values displayed on the figures are adjusted). To test whether the bivariate means (1- richness and 2 - the number of donors from which specific OTUs engrafted) are equal between per visit groups of responses, a multivariate ANOVA (MANOVA) analysis was performed. OS 采用 Kaplan-Meier 法计算。OS定义为从D0到死亡的时间，与病因无关。所有 p 值均为 2 边值，任何 p < 0.05p<0.05 值均被视为具有统计学意义。对于单个健康供体与 MaaT013 批次之间微生物组的比较，我们采用了 Fisher's F 方差齐性检验。我们计算了供体（所有活动）和每个生产活动中 MaaT013 批次的布雷-柯蒂斯相似性指数（1 减布雷-柯蒂斯不相似性指数）。我们使用 Fischer 检验和 Wilcoxon 检验分别测试了五个生产活动中单个健康供体和 MaaT013 批次之间相似性测量的方差均一性和中位数相等性（对于这两种检验，我们根据五个比较的误发现率 (FDR) ^(21){ }^{21} 调整了 p 值）。为了评估单个供体和集合之间丰富度的差异，我们使用了韦尔奇异方差校正 t 检验。在微生物组和粪便代谢物比较方面，为了比较D0、9、16和28时有反应组（R，D28时处于 CR,VGPR\mathrm{CR}, \mathrm{VGPR} 或PR的患者）和无反应组（NR，D28时未处于CR、VGPR或PR的患者），以及比较D0和D9、D0和D16、D0和D28时的R组，我们进行了7次非配对双侧Wilcoxon检验。在移植方面，比较了 RR 组和NR组在D9、16和28时移植的操作分类单元（OTU）的百分比，这些OTU是在治疗前仅在产品中、仅在患者中或同时在产品和患者中确定的。经 FDR 控制多重检验（3 次移植指标检验，7 次微生物组和粪便代谢物比较检验）调整后，当 p 值小于 0 时，则认为 p 值显著。05 （图中显示的 p 值经过调整）。为了检验每次访问各组答复之间的双变量均值（1-丰富度和 2-插入特定 OTU 的供体数量）是否相等，我们进行了多变量方差分析（MANOVA）。
All these analyses were programmed and executed using the “stats” core package of the R software (v 4.2.1). Details on immune and faecal parameter analyses including bioinformatic analyses are provided in the Supplementary file. 所有这些分析都是使用 R 软件（4.2.1 版）的 "stats "核心软件包进行编程和执行的。包括生物信息学分析在内的免疫和粪便参数分析详情见补充文件。
Compassionate use and expanded access program (EAP) 同情使用和扩大使用计划（EAP）
From July 2018 to April 8, 2021, clinical data from 52 GvHD patients who were not eligible for the HERACLES trial and treated through a compassionate use and EAP were collected. The protocol for the EAP is available in the Supplementary Materials. French health authorities (ANSM) authorisation was obtained for each patient before MaaT013 administration. Similarly for HERACLES, a total of 3 MaaT013 doses ( 7+-27 \pm 2 days between each administration) were planned. A “protocol for therapeutic use” validated by ANSM was provided to all 从 2018 年 7 月到 2021 年 4 月 8 日，收集了 52 名不符合 HERACLES 试验条件、通过同情使用和 EAP 接受治疗的 GvHD 患者的临床数据。EAP 协议见补充材料。每位患者在使用 MaaT013 之前都获得了法国卫生当局 (ANSM) 的授权。与 HERACLES 相似，计划总共使用 3 次 MaaT013（每次用药间隔 7+-27 \pm 2 天）。经 ANSM 验证的 "治疗使用协议 "已提供给所有患者。
14 centres to ensure consistency of procedure, clinical management, and data collection (Supplementary file). The EAP population comprised patients with GI-aGvHD with known resistance to, or dependence on CS alone or with failure of other treatment lines used (including the last line of treatment used) for known CS resistance/ dependence. Steroid-resistance definition was the same as in HERACLES. Steroid dependence was defined as the inability to taper CS therapy below 0.5mg//kg//0.5 \mathrm{mg} / \mathrm{kg} / day. Patients with acute GI manifestation of overlap syndrome (features of both aGvHD and chronic GvHD) were eligible. A complete list of eligibility and exclusion criteria is provided in Supplementary Table S4. GI-response at D28 was assessed using the same definitions as in the HERACLES trial. For patients resistant to first-line systemic therapy with CS treated within the EAP, they were not treated within the prospective study because they were treated in a centre not open to inclusion, they did not met inclusion/exclusion criteria or they were treated after the end of study inclusion. 14个中心，以确保程序、临床管理和数据收集的一致性（补充文件）。EAP人群包括已知对CS耐药或依赖CS的GI-aGvHD患者，或因已知CS耐药/依赖而使用其他治疗方法（包括最后一种治疗方法）失败的患者。类固醇耐药性的定义与 HERACLES 相同。类固醇依赖的定义是无法将CS治疗减量至 0.5mg//kg//0.5 \mathrm{mg} / \mathrm{kg} / 天以下。具有重叠综合征急性消化道表现（同时具有急性GvHD和慢性GvHD特征）的患者符合条件。资格和排除标准的完整列表见补充表 S4。采用与 HERACLES 试验相同的定义评估 D28 时的消化道反应。对于在EAP范围内接受CS一线系统治疗的耐药患者，由于他们在不开放纳入的中心接受治疗、不符合纳入/排除标准或在纳入研究结束后接受治疗，因此未在前瞻性研究中接受治疗。
Ethics 伦理学
The HERACLES trial was approved by national ethics committees (CPP Sud Méditerranée V in France on 30 January 2018, West London & GTAC Research Ethics Committee in UK on 31 July 2019, Ethikkommission an der Universität Regensburg in Germany on 07 March 2019, Fondazione Policlinico Universitario Agostino Gemelli IRCCS in Italy on 06 November 2018 and Niezalezna Komisja Bioetyczna ds Badan Naukowych przy GUM in Poland on 18 June 2018). The study was approved by national competent authorities (ANSM in France on 22 December 2017, BfArM in Germany on 11 January 2019, AIFA in Italy on 03 July 2018, MHRA in UK on 09 July 2019 and URPL in Poland on 13 August HERACLES试验获得了各国伦理委员会的批准（法国CPP Sud Méditerranée V，2018年1月30日；英国West London & GTAC研究伦理委员会，2019年7月31日；德国Ethikkommission an der Universität Regensburg，2019年3月7日；意大利Fondazione Policlinico Universitario Agostino Gemelli IRCCS，2018年11月6日；波兰Niezalezna Komisja Bioetyczna ds Badan Naukowych przy GUM，2018年6月18日）。该研究获得了各国主管机构的批准（法国 ANSM 于 2017 年 12 月 22 日批准，德国 BfArM 于 2019 年 1 月 11 日批准，意大利 AIFA 于 2018 年 7 月 3 日批准，英国 MHRA 于 2019 年 7 月 9 日批准，波兰 URPL 于 8 月 13 日批准）。
2018). EudraCT number was 2017-002697-39. For the compassionate use and EAP, patients were treated through the compassionate use program from July 2018 to July 2019 and through a formalised named-patient use program called “Autorisation Temporaire d’Utilisa-tion-ATU nominative” (nominative EAP) from July 24, 2019, with a governing protocol authorised by the ANSM. All patients provided informed consent. 2018).EudraCT 编号为 2017-002697-39。对于同情性使用和EAP，患者在2018年7月至2019年7月期间通过同情性使用计划接受治疗，并从2019年7月24日起通过名为 "Autorisation Temporaire d'Utilisa-tion-ATU nominative"（提名EAP）的正式命名患者使用计划接受治疗，其管理方案由ANSM授权。所有患者均提供了知情同意书。
Role of the funding source 资金来源的作用
MaaT Pharma was the funding source and the legal sponsor of this study and collaborated with the investigators to run the study and helped with the collection and the analysis of data. MaaT Pharma was not involved in the interpretation of data, writing of the report, and the decision to submit the paper for publication. MaaT Pharma 是本研究的资金来源和法律赞助商，与研究人员合作开展研究，并协助收集和分析数据。MaaT Pharma 没有参与数据解读、报告撰写和论文发表的决定。
Results 成果
Between Aug 13, 2018, and Feb 24, 2020, 24 adult patients with SR-GI-aGvHD were enrolled in HERACLES; the last patient’s final visit took place on Nov 26, 2020 (See Fig. 1 for CONSORT diagram and Fig. 2A for study flow chart.). All patients developed a first episode of grade III-IV aGvHD with gut predominance (if any other organ involved) and were resistant to first-line systemic therapy with CS (See Table 1 and Supplementary Table S5 for patients’ characteristics). GI aGvHD stage was 2 in 6 patients (25%), 3 in 9 patients (38%) and 4 in 9 patients ( 38%38 \% ). All patients but 3 had a high-risk Minnesota risk score. ^(22){ }^{22} The median MAGIC algorithm probability (MAP) score ^(23){ }^{23} calculated based on ST2 and REG3 alpha\alpha was 0.993 (range 0.706-10.706-1 ), indicating that all patients were high-risk. ^(23){ }^{23} The median time from 2018年8月13日至2020年2月24日期间，24名SR-GI-aGvHD成人患者入组HERACLES；最后一名患者的最后一次就诊时间为2020年11月26日（CONSORT图见图1，研究流程图见图2A）。所有患者均首次出现以肠道为主的 III-IV 级 aGvHD（如有其他器官受累），且对 CS 的一线系统治疗耐药（患者特征见表 1 和补充表 S5）。6名患者（25%）、9名患者（38%）和9名患者（ 38%38 \% ）的消化道aGvHD分期分别为2期、3期和4期。除3名患者外，其他患者的明尼苏达风险评分均为高风险。 ^(22){ }^{22} 根据ST2和REG3 alpha\alpha 计算出的MAGIC算法概率（MAP）得分 ^(23){ }^{23} 中位数为0.993（范围 0.706-10.706-1 ），表明所有患者均为高危。 ^(23){ }^{23} 中位时间从
All patients received at least 1 MaaT013 dose, 22 ( 92%92 \% ) at least 2 doses, and 12 (50%) received 3 doses (full treatment sequence) (Supplementary Table S7). Reasons for treatment discontinuation were death (n=5)(\mathrm{n}=5), lack of efficacy (n=5)(\mathrm{n}=5), and intensive care unit hospitalisation (n=2)(\mathrm{n}=2). Considering all treated patients ( n=24\mathrm{n}=24 ), 5 achieved a complete resolution of GI-aGvHD symptoms (CR) and 2 a VGPR by D28 after inclusion. Thus, the primary endpoint was met with a combined response of 29%29 \% (CR + VGPR) at D28. Including PR (n=2(\mathrm{n}=2 ), the overall response rate (ORR) was 38%38 \%. When considering GvHD in all organs, the ORR was 33%33 \%. Median duration of response for GI-aGvHD was 164 days (range, 4-352) (Fig. 2B): at last follow-up, 4 patients were still in CR, 3 were still in CR but died from haematological malignancy relapse ( n=2\mathrm{n}=2 ) and pulmonary infection (n=1(\mathrm{n}=1 ), and 2 had GvHD relapse (following antibiotic treatment for one patient). 所有患者至少接受了 1 次 MaaT013 治疗，22 人（ 92%92 \% ）至少接受了 2 次治疗，12 人（50%）接受了 3 次治疗（完整治疗序列）（补充表 S7）。停止治疗的原因是死亡 (n=5)(\mathrm{n}=5) 、疗效不佳 (n=5)(\mathrm{n}=5) 和重症监护室住院 (n=2)(\mathrm{n}=2) 。考虑到所有接受治疗的患者（ n=24\mathrm{n}=24 ），5名患者在入组后的第28天实现了消化道-aGvHD症状的完全缓解（CR），2名患者实现了VGPR。因此，在 D28 时， 29%29 \% （CR + VGPR）的综合反应达到了主要终点。包括 PR (n=2(\mathrm{n}=2 ），总反应率（ORR）为 38%38 \% 。如果考虑所有器官的 GvHD，ORR 为 33%33 \% 。GI-aGvHD 反应持续时间的中位数为 164 天（4-352 天不等）（图 2B）：最后一次随访时，4 名患者仍处于 CR 期，3 名患者仍处于 CR 期，但死于血液恶性肿瘤复发（ n=2\mathrm{n}=2 ）和肺部感染（ (n=1(\mathrm{n}=1 ），2 名患者 GvHD 复发（1 名患者经抗生素治疗后复发）。
The OS rate was 67%67 \% ( 95%95 \% CI, 44-82) after 30 days, 28%28 \% ( 95%95 \% CI, 13-47) at 6 months and 25%25 \% ( 95%95 \% CI, 10-43) at one year (Fig. 2C). Median survival duration was 91 days [ 95%95 \% CI 25 ; 171]. The cause of death was GvHD in 8 patients, severe infection in 6 , relapse of underlying malignancy in 2 , cardiac failure in 1 and respiratory failure in 1. 30天后的OS率为 67%67 \% （ 95%95 \% CI，44-82），6个月后的OS率为 28%28 \% （ 95%95 \% CI，13-47），一年后的OS率为 25%25 \% （ 95%95 \% CI，10-43）（图2C）。中位生存期为91天[ 95%95 \% CI 25 ; 171]。8例患者的死因是急性缺血性坏死，6例是严重感染，2例是基础恶性肿瘤复发，1例是心力衰竭，1例是呼吸衰竭。
Most frequent AEs and SAEs are provided in Table 2. Thirty-nine AEs including 4 SAEs occurred within 24 h of MaaT013 dose administration (Supplementary Table S8). Among them, 16 occurred after the first administration, including 1 SAE (life-threatening cerebral infarction). Sixteen happened after the second administration, including 3 SAEs (thrombotic microangiopathy, Escherichia coli (E. coli) sepsis, and fatal general physical health deterioration). Finally, 7 AEs occurred after the third administration of MaaT013, of which none were considered to be serious. No death occurred within 24 h after MaaT013 administration. In addition, beyond 24 h of MaaT013 dose administration, 4 SAEs possibly related to MaaT013 (E. coli urinary tract infection, enterococcal urinary tract infection, E. coli sepsis and Staphylococcus epidermidis sepsis) and one SAE probably related to 表 2 列出了最常见的 AE 和 SAE。39 例 AE（包括 4 例 SAE）发生在 MaaT013 服药后 24 小时内（补充表 S8）。其中，16 例发生在首次给药后，包括 1 例 SAE（危及生命的脑梗塞）。16例发生在第二次给药后，包括3例SAE（血栓性微血管病、大肠杆菌败血症和致命性全身健康恶化）。最后，在第三次服用MaaT013后发生了7例AE，其中无一例被认为是严重的。MaaT013用药后24小时内未出现死亡病例。此外，在服用 MaaT013 24 小时后，发生了 4 例可能与 MaaT013 有关的 SAE（大肠杆菌尿路感染、肠球菌尿路感染、大肠杆菌败血症和表皮葡萄球菌败血症）和 1 例可能与 MaaT013 有关的 SAE（大肠杆菌尿路感染、肠球菌尿路感染、大肠杆菌败血症和表皮葡萄球菌败血症）。
MaaT013 (E. coli sepsis) were reported, none of which led to the patient’s death. For all events, shotgun sequencing analyses of the identified strains revealed that none were found in MaaT013 (detection limit: 0.3%0.3 \% relative abundance of the strain in the product). MaaT013（大肠杆菌败血症），但均未导致患者死亡。对所有事件中已确定的菌株进行枪式测序分析后发现，MaaT013 中未发现任何菌株（检测限： 0.3%0.3 \% 产品中菌株的相对丰度）。
Bacterial phylogenetic composition was determined in HERACLES patients using 16S rDNA sequencing and alpha\alpha - and beta\beta-diversity indices were calculated for faecal samples (rectal swabs), collected from patients at baseline (D0, before first MaaT013 dose), D9 (before 2nd dose), D16 (before 3rd dose) and D28 when treatment response was evaluated. Using LEfSe, we found differences between the gut microbiota composition of R versus NR patients at baseline and after treatment (Supplementary result and Supplementary Figs. S2-S4). 使用 16S rDNA 测序确定了 HERACLES 患者的细菌系统发育组成，并计算了粪便样本（直肠拭子）的 alpha\alpha - 和 beta\beta - 多样性指数，粪便样本采集于基线（D0，首次服用 MaaT013 前）、D9（第二次服用前）、D16（第三次服用前）和评估治疗反应的 D28。通过使用 LEfSe，我们发现在基线和治疗后，R 患者与 NR 患者的肠道微生物群组成存在差异（补充结果和补充图 S2-S4）。
We also found that bacteria previously associated with improved outcomes after allo-HCT were associated with response. The Butycore, a sum of relative abundances of a set of butyrate-producing genera, was significantly higher in R versus NR patients, as early as D9 (after first MaaT013 administration), and up to D28 (at visit D9 median 14.6%14.6 \% versus 0.2%,p=0.0440.2 \%, \mathrm{p}=0.044, at D16 13.0%13.0 \% versus 0.1%,p=0.0440.1 \%, \mathrm{p}=0.044, and at D28 10.1%10.1 \% versus 0.7%,p=0.0410.7 \%, p=0.041 Fig. 3A). In RR patients, the Butycore was significantly increased at D9, D16 and D28 compared to baseline ( p=0.019p=0.019 at each timepoint). Furthermore, butyrate producers, Clostridiales and Blautia were also significantly increased in R at D9, D16, and D28 compared to baseline (Fig. 3B and C). 我们还发现，以前与异体肝细胞移植后预后改善有关的细菌也与反应有关。Butycore是一组丁酸盐产生菌属的相对丰度总和，早在D9（首次给予MaaT013后）和直到D28（D9中位数 14.6%14.6 \% 对 0.2%,p=0.0440.2 \%, \mathrm{p}=0.044 ，D16 13.0%13.0 \% 对 0.1%,p=0.0440.1 \%, \mathrm{p}=0.044 ，D28 10.1%10.1 \% 对 0.7%,p=0.0410.7 \%, p=0.041 图3A），R患者的Butycore明显高于NR患者。在 RR 患者中，与基线相比，丁酸核在D9、D16和D28显著增加（每个时间点均为 p=0.019p=0.019 ）。此外，与基线相比，丁酸盐生产者、梭杆菌和布劳氏菌在 D9、D16 和 D28 也在 R 中明显增加（图 3B 和 C）。
When looking at faecal metabolites, we found that SCFA butyrate and isobutyrate levels were increased, although the increases were not statistically significant, in R compared to NR patients at D9 after the first dose (median 9.67 mug//g9.67 \mu \mathrm{~g} / \mathrm{g} of faeces versus 0.01 mug//g,p=0.190.01 \mu \mathrm{~g} / \mathrm{g}, \mathrm{p}=0.19 for butyrate and 2.50 mug//g2.50 \mu \mathrm{~g} / \mathrm{g} versus 0.06 mug//g,p=0.0630.06 \mu \mathrm{~g} / \mathrm{g}, \mathrm{p}=0.063 for isobutyrate) (Fig. 3D and E). The same observation was made on faecal bile secondary acids, with significantly increased levels at D9 after the first MaaT013 administration in R compared to NR patients (Supplementary Fig. S5). Finally, we assessed plasma levels of 3-indoxyl sulfate (3-IS), a metabolite associated with a disrupted microbiota and poor outcomes after allo-HCT. ^(24){ }^{24} We found that while there was no difference in 3-IS level at baseline, it increased, although this increase was not statistically significant, in R patients compared to baseline at D9 and D28 ( p=0.057p=0.057 and p=0.057p=0.057, respectively) and higher levels were achieved compared to NR patients at D9 (median 3.2 versus 1.4 mumol//L,p=0.0571.4 \mu \mathrm{~mol} / \mathrm{L}, \mathrm{p}=0.057 ) (Fig. 3F). 在观察粪便代谢物时，我们发现与 NR 患者相比，R 患者的 SCFA 丁酸盐和异丁酸盐水平在首次给药后的第 9 天有所增加，但增加幅度在统计学上并不显著（粪便中位数 9.67 mug//g9.67 \mu \mathrm{~g} / \mathrm{g} 与 0.01 mug//g,p=0.190.01 \mu \mathrm{~g} / \mathrm{g}, \mathrm{p}=0.19 相比，丁酸盐和异丁酸盐分别为 2.50 mug//g2.50 \mu \mathrm{~g} / \mathrm{g} 与 0.06 mug//g,p=0.0630.06 \mu \mathrm{~g} / \mathrm{g}, \mathrm{p}=0.063 ）（图 3D 和 E）。对粪便胆汁仲酸也进行了同样的观察，与 NR 患者相比，R 患者在首次服用 MaaT013 后的第 9 天，粪便胆汁仲酸水平显著升高（补充图 S5）。最后，我们评估了血浆中3-吲哚硫酸酯（3-IS）的水平，这是一种与异体HCT后微生物群紊乱和不良预后相关的代谢物。 ^(24){ }^{24} 我们发现，虽然3-IS水平在基线时没有差异，但在D9和D28时，R患者的3-IS水平与基线时相比有所增加，尽管这种增加没有统计学意义（分别为 p=0.057p=0.057 和 p=0.057p=0.057 ），而且在D9时，R患者的3-IS水平比NR患者更高（中位数为3.2比 1.4 mumol//L,p=0.0571.4 \mu \mathrm{~mol} / \mathrm{L}, \mathrm{p}=0.057 ）（图3F）。
In addition, alpha-diversity was evaluated using Shannon and richness indices at the OTU level 此外，还在 OTU 一级使用香农指数和丰富度指数对阿尔法多样性进行了评估
Treated patients 接受治疗的患者(n=24)(n=24)
Sex, n (%) 性别，n (%)
Male 男
16 (67%)
Female 女性
8 (33%)
Age at inclusion (years) 纳入时的年龄（岁）
Median [range] 中位数 [范围］
61[20;69]61[20 ; 69]
Diagnosis, n (%) 诊断，n (%)
Acute myeloid leukaemia 急性髓性白血病
9 (38%)
Acute lymphoblastic leukaemia 急性淋巴细胞白血病
2(8%)2(8 \%)
Myelodysplastic syndrome 骨髓增生异常综合征
3 (13%)
Myeloproliferative syndrome 骨髓增生异常综合征
5 (21%)
Chronic myeloid leukaemia 慢性骨髓性白血病
1 (4%)
Hodgkin lymphoma 霍奇金淋巴瘤
1 (4%)
Other 其他
3(13%)3(13 \%)
Graft source, n (%) 移植物来源，n (%)
Bone marrow 骨髓
1 (4%)
Peripheral blood stem cells 外周血干细胞
23(96%)23(96 \%)
Type of donor, n(%)\mathrm{n}(\%) 捐赠者类型， n(%)\mathrm{n}(\%)
Related 相关
11(46%)11(46 \%)
Unrelated 无关
13(54%)13(54 \%)
Status of the disease before allo-HCT, n (%) 异体肝细胞移植前的疾病状况，n (%)
Initial diagnosis 初步诊断
4 (17%)
Complete remission 完全缓解
8 (33%)
Partial remission 部分缓解
3 (13%)
Refractory or progression 难治或病情恶化
9 (38%)
Conditioning regimen 调理方案
Non myeloablative 非骨髓溶解
4 (17%)
Reduced intensity 降低强度
8 (33%)
Reduced toxicity 降低毒性
12(50%)12(50 \%)
In vivo T-cell depletion, n(%)\mathrm{n}(\%) 体内 T 细胞耗竭、 n(%)\mathrm{n}(\%)
Time from allo-HCT date to aGvHD onset (days) 从异体肝细胞移植日期到发生急性肾脏病的时间（天数）
Median [range] 中位数 [范围］
34[12;207]34[12 ; 207]
Time from start of CS therapy to inclusion (days) 从开始 CS 治疗到纳入的时间（天数）
Median [range] 中位数 [范围］
8.5[4.0;48.0]8.5[4.0 ; 48.0]
Total dose of CS (methylprednisolone-equivalent) per day at inclusion ( mg//kg\mathrm{mg} / \mathrm{kg} ) 纳入时每天 CS 的总剂量（甲基强的松龙当量） ( mg//kg\mathrm{mg} / \mathrm{kg} )
Median [range] 中位数 [范围］
2.0[1.2;2.2]2.0[1.2 ; 2.2]
Time from start of CS therapy to (days) 从开始 CS 治疗到（天）的时间
Median [range] 中位数 [范围］
7.0[4.0;34.0]7.0[4.0 ; 34.0]
Time from SR diagnosis to inclusion (days), median [min; max] 从 SR 诊断到纳入的时间（天），中位数[最小；最大］
Median [range] 中位数 [范围］
2.0[0;14.0]2.0[0 ; 14.0]
Minnesota risk score 明尼苏达州风险评分
Standard risk 标准风险
3(13%)3(13 \%)
High risk 高风险
21(87%)21(87 \%)
Treated patients (n=24)
Sex, n (%)
Male 16 (67%)
Female 8 (33%)
Age at inclusion (years)
Median [range] 61[20;69]
Diagnosis, n (%)
Acute myeloid leukaemia 9 (38%)
Acute lymphoblastic leukaemia 2(8%)
Myelodysplastic syndrome 3 (13%)
Myeloproliferative syndrome 5 (21%)
Chronic myeloid leukaemia 1 (4%)
Hodgkin lymphoma 1 (4%)
Other 3(13%)
Graft source, n (%)
Bone marrow 1 (4%)
Peripheral blood stem cells 23(96%)
Type of donor, n(%)
Related 11(46%)
Unrelated 13(54%)
Status of the disease before allo-HCT, n (%)
Initial diagnosis 4 (17%)
Complete remission 8 (33%)
Partial remission 3 (13%)
Refractory or progression 9 (38%)
Conditioning regimen
Non myeloablative 4 (17%)
Reduced intensity 8 (33%)
Reduced toxicity 12(50%)
In vivo T-cell depletion, n(%)
ATG alone 11(46%)
PTCy alone 6 (25%)
ATG + PTCy 1(4%)
None 6(25%)
GvHD prophylaxis, n (%)
CsA alone ^(a) 7 (29%)
CsA+MMF^("b ") 12 (50%)
CSA + MTX 4 (17%)
MMF+CS 1(4%)
Time from allo-HCT date to aGvHD onset (days)
Median [range] 34[12;207]
Time from start of CS therapy to inclusion (days)
Median [range] 8.5[4.0;48.0]
Total dose of CS (methylprednisolone-equivalent) per day at inclusion ( mg//kg )
Median [range] 2.0[1.2;2.2]
Time from start of CS therapy to (days)
Median [range] 7.0[4.0;34.0]
Time from SR diagnosis to inclusion (days), median [min; max]
Median [range] 2.0[0;14.0]
Minnesota risk score
Standard risk 3(13%)
High risk 21(87%)| | Treated patients $(n=24)$ |
| :---: | :---: |
| Sex, n (%) | |
| Male | 16 (67%) |
| Female | 8 (33%) |
| Age at inclusion (years) | |
| Median [range] | $61[20 ; 69]$ |
| Diagnosis, n (%) | |
| Acute myeloid leukaemia | 9 (38%) |
| Acute lymphoblastic leukaemia | $2(8 \%)$ |
| Myelodysplastic syndrome | 3 (13%) |
| Myeloproliferative syndrome | 5 (21%) |
| Chronic myeloid leukaemia | 1 (4%) |
| Hodgkin lymphoma | 1 (4%) |
| Other | $3(13 \%)$ |
| Graft source, n (%) | |
| Bone marrow | 1 (4%) |
| Peripheral blood stem cells | $23(96 \%)$ |
| Type of donor, $\mathrm{n}(\%)$ | |
| Related | $11(46 \%)$ |
| Unrelated | $13(54 \%)$ |
| Status of the disease before allo-HCT, n (%) | |
| Initial diagnosis | 4 (17%) |
| Complete remission | 8 (33%) |
| Partial remission | 3 (13%) |
| Refractory or progression | 9 (38%) |
| Conditioning regimen | |
| Non myeloablative | 4 (17%) |
| Reduced intensity | 8 (33%) |
| Reduced toxicity | $12(50 \%)$ |
| In vivo T-cell depletion, $\mathrm{n}(\%)$ | |
| ATG alone | $11(46 \%)$ |
| PTCy alone | 6 (25%) |
| ATG + PTCy | $1(4 \%)$ |
| None | $6(25 \%)$ |
| GvHD prophylaxis, n (%) | |
| CsA alone ${ }^{a}$ | 7 (29%) |
| $\mathrm{CsA}+\mathrm{MMF}^{\text {b }}$ | 12 (50%) |
| CSA + MTX | 4 (17%) |
| $M M F+C S$ | $1(4 \%)$ |
| Time from allo-HCT date to aGvHD onset (days) | |
| Median [range] | $34[12 ; 207]$ |
| Time from start of CS therapy to inclusion (days) | |
| Median [range] | $8.5[4.0 ; 48.0]$ |
| Total dose of CS (methylprednisolone-equivalent) per day at inclusion ( $\mathrm{mg} / \mathrm{kg}$ ) | |
| Median [range] | $2.0[1.2 ; 2.2]$ |
| Time from start of CS therapy to (days) | |
| Median [range] | $7.0[4.0 ; 34.0]$ |
| Time from SR diagnosis to inclusion (days), median [min; max] | |
| Median [range] | $2.0[0 ; 14.0]$ |
| Minnesota risk score | |
| Standard risk | $3(13 \%)$ |
| High risk | $21(87 \%)$ |
(Supplementary Table S9). The mean Shannon index (all patients) increased progressively from 1.4 to 2.3 at D28. Changes in the Shannon index when compared to baseline became significantly higher in RR as early as D9D 9 (D9 versus D0, p=0.020\mathrm{p}=0.020 and D28 versus D0, p=0.004\mathrm{p}=0.004 ) (Fig. 4A). Mean Richness index (all patients) increased from 46.8 to 107.6 at D28. Changes in richness index when compared to baseline were significant in R patients at every time point ( D 9 versus D0,p=0.012\mathrm{D} 0, \mathrm{p}=0.012; D16 versus D0, p=0.025\mathrm{p}=0.025; D28 versus D0, p=0.012\mathrm{p}=0.012 ) (Fig. 4B). Beta-diversity was evaluated using the Bray-Curtis similarity index (calculated as 1 minus the BrayCurtis dissimilarity index), when comparing patients’ gut microbiota composition to MaaT013 composition. The mean Bray-Curtis similarity index increased progressively from 0.04 to 0.16 at D28, and it increased significantly in RR patients compared to baseline (补充表 S9）。平均香农指数（所有患者）从 1.4 逐步上升到 D28 时的 2.3。与基线相比， RR 的香农指数变化早在 D9D 9 时就明显变大（D9与D0相比， p=0.020\mathrm{p}=0.020 ；D28与D0相比， p=0.004\mathrm{p}=0.004 ）（图 4A）。平均丰富度指数（所有患者）从 46.8 增加到 D28 时的 107.6。与基线相比，R 患者的丰富度指数在每个时间点都有显著变化（D 9 与 D0,p=0.012\mathrm{D} 0, \mathrm{p}=0.012 ；D16 与 D0， p=0.025\mathrm{p}=0.025 ；D28 与 D0， p=0.012\mathrm{p}=0.012 ）（图 4B）。在比较患者肠道微生物群组成与 MaaT013 组成时，使用布雷-柯蒂斯相似性指数（计算方法为 1 减布雷-柯蒂斯不相似性指数）来评估 Beta 多样性。平均布雷-柯蒂斯相似性指数从 0.04 逐步增加到 D28 时的 0.16，与基线相比， RR 患者的相似性指数显著增加
(MaaT013-D9 versus MaaT013-D0, p=0.013\mathrm{p}=0.013; MaaT013D16 versus MaaT013-D0, p=0.026\mathrm{p}=0.026; MaaT013-D28 versus MaaT013-D0, p=0.026\mathrm{p}=0.026, Fig. 4C). The BrayCurtis similarity index was not significantly different when comparing RR and NR patients. (MaaT013-D9与MaaT013-D0相比， p=0.013\mathrm{p}=0.013 ；MaaT013D16与MaaT013-D0相比， p=0.026\mathrm{p}=0.026 ；MaaT013-D28与MaaT013-D0相比， p=0.026\mathrm{p}=0.026 ，图4C）。在比较 RR 和NR患者时，BrayCurtis相似性指数没有明显差异。
We then looked at the engraftment of MaaT013 in the patients’ gut and found that the proportion of OTUs in the stool coming from MaaT013 was higher in R than in NR patients at D28, however, it did not reach statistical significance (median 58%58 \% versus 26%,p=0.1326 \%, p=0.13 ). In contrast, NR patients had significantly more OTUs of patient origin at D28 ( p=0.024\mathrm{p}=0.024, Fig. 4D and Supplementary Fig. S6). 然后，我们观察了 MaaT013 在患者肠道中的移植情况，发现在第 28 天时，R 患者粪便中来自 MaaT013 的 OTU 比例高于 NR 患者，但两者没有达到统计学意义（中位数 58%58 \% 对 26%,p=0.1326 \%, p=0.13 ）。相比之下，NR患者在D28时患者来源的OTU明显增多（ p=0.024\mathrm{p}=0.024 ，图4D和补充图S6）。
The relationship between patient microbiota richness at each visit and the number of donors whose specific material contained within MaaT013 engrafted into the patients’ gut was investigated (Fig. 4E). A multivariate analysis of these 2 parameters (richness and number of donors) could significantly distinguish between RR and NR populations (MANOVA, p < 0.001\mathrm{p}<0.001, clinical response at each visit). R patients displayed a high level of richness correlating with the presence of microbial communities from multiple donors (median =2=2 ), whereas NR patients had engrafted material from fewer donors (median =0.5=0.5 ), strongly suggesting that the clinical GI response to MaaT013, at any point in time, is linked in most patients, to the engraftment of microbial communities from multiple donors ( >= 2)(\geq 2). Similarly (Supplementary Fig. S7), the engraftment percentage increases with the number of donors who contributed to the received MaaT013, and both variables are indicative of the response per visit (MANOVA, p < 0.001\mathrm{p}<0.001 ). 研究了每次就诊时患者微生物群丰富度与 MaaT013 中所含特异性物质移植到患者肠道的供体数量之间的关系（图 4E）。对这两个参数（丰富度和供体数量）的多变量分析可以显著区分 RR 和 NR 群体（MANOVA， p < 0.001\mathrm{p}<0.001 ，每次就诊时的临床反应）。R型患者显示出较高的丰富度，这与多个供体的微生物群落（中位数 =2=2 ）的存在有关，而NR型患者的移植材料来自较少的供体（中位数 =0.5=0.5 ），这有力地表明，在任何时间点，大多数患者对MaaT013的临床消化道反应都与多个供体的微生物群落的移植有关 ( >= 2)(\geq 2) 。同样（补充图 S7），随着接受 MaaT013 的供体数量的增加，移植百分比也在增加，这两个变量都表明了每次就诊的反应（MANOVA， p < 0.001\mathrm{p}<0.001 ）。
Principal component analysis (PCA) performed on host parameters and the microbiota shows the projections of samples from patients before their first baseline (BL) treatments and after treatment with their per-visit GI response classification ( RR and NRN R ) (Supplementary Fig. S8 and S9). For host parameters (Supplementary Fig. S8), R point projections are mainly explained by an increase in SCFAs, deoxy- and lithocholic acids, 3IS, TGFb 1 and 2, CCL25 and citrulline and lower levels of CRP, interleukins (IL-6, IL-8, IL-2, IL1b, and IL-18), neopterin (blood), ferritin, CCL2/MCP-1, sCD30 and Total Antioxidant Status compared to BL and NR measurement. For the microbiota (Supplementary Fig. S9): R point projections are mainly explained by an increase in Phascolarctobacterium, Alistipes, Desulfovibrio, Butyricicoccus, Collinsella, Lachnospiraceae UCG-08, Subdoligranulum, Bilophila, Faecalibacterium, Dialister, and Blautia for the main genera, and lower levels of Corynebacterium, Staphylococcus, Enterococcus, Lactobacillus as well as an unassigned genus, compared to the BL or NR measurements. 对宿主参数和微生物群进行的主成分分析（PCA）显示了患者样本在首次基线（BL）治疗前和治疗后与每次就诊消化道反应分类（ RR 和 NRN R ）的投影（补充图 S8 和 S9）。对于宿主参数（补充图 S8），与 BL 和 NR 测量相比，R 点预测的主要原因是 SCFAs、脱氧胆酸和石胆酸、3IS、TGFb 1 和 2、CCL25 和瓜氨酸的增加，以及 CRP、白细胞介素（IL-6、IL-8、IL-2、IL1b 和 IL-18）、蝶呤（血）、铁蛋白、CCL2/MCP-1、sCD30 和总抗氧化状态水平的降低。对于微生物群（补充图 S9）：R 点的预测主要是由于在 BL 和 NR 测量中，Phascolarctobacterium、Alistipes、Desulfovibrio、Butyricicoccus、Collinsella、Lachnospiraceae UCG-08、Subdoligranulum、Bilophila、Faecalibacterium、Dialister 和 Blautia 的增加、与 BL 或 NR 测量值相比，主要菌属中的棒状杆菌、葡萄球菌、肠球菌、乳酸杆菌以及一 个未确定的菌属含量较低。
Both host parameters and faecal microbiota analyses were reconciled on a common factorial map through the co-inertia analysis (CIA) procedure (Fig. 5). In Fig. 5 A , B,R-NR-BL\mathrm{B}, \mathrm{R}-\mathrm{NR}-\mathrm{BL} clusters are consistent, meaning that both biomarkers and the microbiota can explain the whole sample variability and are both able to separate the clusters (the co-structure strength as explained with the RV coefficient is significant with p=0.001\mathrm{p}=0.001, Fig. 5E). 通过共惯性分析（CIA）程序（图 5），宿主参数和粪便微生物群分析在一个共同的因子图上得到了协调。在图 5 A 中， B,R-NR-BL\mathrm{B}, \mathrm{R}-\mathrm{NR}-\mathrm{BL} 聚类是一致的，这意味着生物标记物和微生物群都能解释整个样本的变异性，并且都能分离聚类（用 RV 系数解释的共结构强度与 p=0.001\mathrm{p}=0.001 显著，图 5E）。
It is possible to distinguish 3 groups of host parameters and genera that “co”-increase according to specific response related clusters (Fig. 5C and D). The group composed of SCFAs, bile acids, and 3IS explains the variability in the same way as the genera group composed of Phascolarctobacterium, Alistipes, Desulfovibrio, Butyricicoccus, Collinsella, Lachnospiraceae UCG-008, Subdoligranulum, Bilophila, Faecalibacterium, Dialister, and Blautia and are related to the R patient cluster; the group constituted by CRP, interleukins (IL6, IL-8, IL-2, IL-1b, and IL-18), neopterin (blood), 根据与特定反应相关的群组（图 5C 和 D），可以区分出 3 组 "共同 "增加的宿主参数和属种。由 SCFAs、胆汁酸和 3IS 组成的组与由 Phascolarctobacterium、Alistipes、Desulfovibrio、Butyricicoccus、Collinsella、Lachnospiraceae UCG-008、Subdoligranulum、Bilophila、Faecalibacterium、Dialister 和 Blautia 组成的属组以相同的方式解释了变异性，并与 R 患者群相关；由 CRP、白细胞介素（IL6、IL-8、IL-2、IL-1b 和 IL-18）、新蝶呤（血液）构成的组、
ferritin, CCL2/MCP-1, and Total Antioxidant Status explains the variability in the same way as the Corynebacterium, Staphylococcus and Enterococcus groups and are related to the NR patient cluster; a third and broader group mainly structured by the second component includes cholesterol and TGFb1 for the host biomarkers, and Actinomyces, Lactobacillus, and Streptococcus for the genera, and tends to explain the distinction between the BL and the NR related clusters. 铁蛋白、CCL2/MCP-1 和总抗氧化状态以与棒状杆菌、葡萄球菌和肠球菌组相同的方式解释了变异性，并与 NR 患者群相关；第三组范围更广，主要由第二部分构成，包括宿主生物标志物中的胆固醇和 TGFb1，以及属中的放线菌、乳酸杆菌和链球菌，并倾向于解释 BL 和 NR 相关群之间的区别。
Shotgun sequencing was performed to evaluate gene functionality using per visit GI-response classification. 利用每次就诊的胃肠道反应分类法进行枪式测序，以评估基因功能。
SCFA-associated Kyoto Encyclopedia of Genes and Genomes (KEGG) Orthology (KO) abundances are significantly increased at D9 in R patients when compared to BL, and this is true for all SCFAs (p=0.029)(p=0.029), and for propionate individually ( p=0.034\mathrm{p}=0.034 ) (Supplementary Fig. S10). Increased KO abundances were observed in R versus NR patients at D16 for all SCFAs and propionate. At D28, no statistical difference was observed between R and NR populations. This difference may be due to the salvage therapy that the 4 NR patients received between D16 and D28 which may have contributed to the restoration of the function of their microbiota, thereby increasing the KO abundance values. 与 BL 相比，R 患者中 SCFA 相关的《京都基因组百科全书》（KEGG）选集（KO）丰度在 D9 时显著增加，所有 SCFA (p=0.029)(p=0.029) 和丙酸盐（ p=0.034\mathrm{p}=0.034 ）均是如此（补充图 S10）。在D16时，观察到R患者与NR患者在所有SCFAs和丙酸酯方面的KO丰度增加。在 D28 时，在 R 和 NR 群体中未观察到统计学差异。这种差异可能是由于 4 名 NR 患者在 D16 和 D28 之间接受了挽救治疗，这可能有助于恢复其微生物群的功能，从而增加了 KO 丰度值。
In addition, clinical data were available for 52 patients with SR or SD-GI-aGvHD treated with MaaT013 as part of a compassionate use program (n=11)(\mathrm{n}=11) and EAP ( n=41\mathrm{n}=41 ) (Table 3 for patients’ characteristics). aGvHD grade was II in 3 patients ( 6%6 \% ) and III in 49 patients (94%). All patients had GI involvement and GIaGvHD stage was 1 in 4 patients ( 8%8 \% ), 2 in 12 patients (23%), 3 in 17 patients ( 33%33 \% ) and 4 in 19 patients ( 37%37 \% ). The median time from aGvHD diagnosis to first administration of MaaT013 was 88 days (range, 11-1337). These patients had previously received and failed 1 to 6 lines (median 3; 40 out of 52 received ruxolitinib) of systemic GvHD treatments. All patients were in failure to their last line of treatment per eligibility criteria. In particular for the 40 patients that received ruxolitinib, 22 patients received additional treatment before MaaT013 administration, while 18 patients received MaaT013 just after ruxolitinib, at a median of 31 (range, 16-613) days, confirming ruxolitinib refractoriness in those patients. ^(25){ }^{25} Among all treated patients (n=52),17(\mathrm{n}=52), 17 achieved a CR on D28 and 9 a VGPR. Including PR (n=4)(\mathrm{n}=4), the GI-ORR was 58%58 \%. In the 9 patients with SD-aGvHD, the D28 GI-ORR was 100%100 \%, including 7 CR and 2 VGPR, and in the 43 patients with SR-aGvHD, the D28 GI-ORR was 49%49 \%, including 10CR,710 \mathrm{CR}, 7 VGPR and 4 PR. In the subgroup of 40 patients who received ruxolitinib (median lines of treatment: 3, range 2-6), the D28-GI-ORR was 58%58 \% including 13CR,713 \mathrm{CR}, 7 VGPR and 3 PR . 此外，作为同情使用计划 (n=11)(\mathrm{n}=11) 和EAP（ n=41\mathrm{n}=41 ）的一部分，有52例接受MaaT013治疗的SR或SD-GI-aGvHD患者的临床数据（患者特征见表3）。所有患者均有消化道受累，消化道aGvHD分期为1期的患者有4例（ 8%8 \% ），2期的患者有12例（23%），3期的患者有17例（ 33%33 \% ），4期的患者有19例（ 37%37 \% ）。从确诊aGvHD到首次使用MaaT013的中位时间为88天（11-1337天）。这些患者之前曾接受过1至6种（中位数为3种；52人中有40人接受了鲁索利替尼治疗）全身性GvHD治疗，但均告失败。根据资格标准，所有患者的最后一线治疗均告失败。特别是在接受鲁索利替尼治疗的40名患者中，22名患者在服用MaaT013前接受了额外治疗，而18名患者是在接受鲁索利替尼治疗后才服用MaaT013，中位数为31天（16-613天），证实了这些患者对鲁索利替尼的耐药。 ^(25){ }^{25} 在所有接受治疗的患者中， (n=52),17(\mathrm{n}=52), 17 在D28达到CR，9例达到VGPR。包括 PR (n=4)(\mathrm{n}=4) 在内，GI-ORR 为 58%58 \% 。在9例SD-aGvHD患者中，D28 GI-ORR为 100%100 \% ，包括7例CR和2例VGPR；在43例SR-aGvHD患者中，D28 GI-ORR为 49%49 \% ，包括 10CR,710 \mathrm{CR}, 7 VGPR和4例PR。在接受鲁索利替尼治疗的 40 例患者亚组（中位治疗线：3，范围 2-6）中，D28 GI-ORR 为 58%58 \% ，包括 13CR,713 \mathrm{CR}, 7 VGPR 和 3 PR。
The median follow-up among surviving patients was 381 days (range, 63-654). The OS rate was 49%49 \% ( 95%95 \% 存活患者的中位随访时间为 381 天（63-654 天不等）。OS 率为 49%49 \% ( 95%95 \% )
CI, 37-64) at 6 months and 38%38 \% ( 95%95 \% CI, 25-52) at 1 year (Supplementary Fig. S11). Median survival duration was 176 days ( 95%95 \% CI, 66-439). The cause of death was GvHD in 15 patients, severe infection in 8, relapse of underlying malignancy in 5, COVID-19 in 3, haemorrhage during surgery in one, and cardiac arrest in one patient. CI，37-64），1年后为 38%38 \% ( 95%95 \% CI，25-52)（补充图 S11）。中位生存期为176天（ 95%95 \% CI，66-439）。15例患者的死因是脑血管疾病，8例是严重感染，5例是基础恶性肿瘤复发，3例是COVID-19，1例是手术中大出血，1例是心脏骤停。
Eighteen pharmacovigilance events were reported for 18 patients and included bacteriemia/sepsis in 11 patients (E. coli, n=5\mathrm{n}=5; Bacteroides fragilis, n=1\mathrm{n}=1; S. epidermidis of central venous catheter origin, n=2\mathrm{n}=2; unknown, n=3\mathrm{n}=3 ); Clostridium difficile colitis in 2, E. coli osteoarthritis in one, presence of Geotrichum silvicola in stools from one, appearance of air bubbles in the mesorectum following MaaT013 administration possibly related to rectal probe trauma in one, and respiratory distress in one patient (see complete list of pharmacovigilance reporting in Supplementary Table S10). 18名患者报告了18起药物警戒事件，包括11名患者的细菌血症/败血症（大肠杆菌， n=5\mathrm{n}=5 ；脆弱拟杆菌， n=1\mathrm{n}=1 ；中心静脉导管来源的表皮葡萄球菌， n=2\mathrm{n}=2 ；未知， n=3\mathrm{n}=3 ）；2名患者的艰难梭菌性结肠炎，1名患者的大肠杆菌性骨关节炎，1名患者的粪便中出现矽藻，MaaT013用药后直肠中膜出现气泡，可能与直肠炎有关。一名患者出现大肠杆菌性骨关节炎，一名患者粪便中出现硅藻土癣菌，一名患者服用 MaaT013 后直肠中膜出现气泡，可能与直肠探针创伤有关，一名患者出现呼吸困难（参见补充表 S10 中的药物警戒报告完整列表）。
Discussion 讨论
Our study demonstrates the feasibility of producing standardised pooled allogeneic faecal microbiota therapy with a higher richness compared to individual healthy donor samples and a standardisation of potential beneficial bacteria such as butyrate-producing bacteria. This standardised pool faecal microbiota MaaT013 translates into a high response rate in patients with SR-GIaGvHD, being 38%38 \% in the phase II HERACLES study and 58%58 \% in the EAP at D28. 我们的研究证明了生产标准化集合异体粪便微生物群疗法的可行性，与单个健康捐献者样本相比，这种疗法具有更高的丰富度，并实现了潜在有益菌（如丁酸菌）的标准化。这种标准化的集合粪便微生物群MaaT013可转化为SR-GIaGvHD患者的高应答率，在II期HERACLES研究中为 38%38 \% ，在D28期EAP中为 58%58 \% 。
Importantly, the EAP included 9 patients with SD-GI-aGvHD that all achieved response at D28 contributing to the higher response rate in these patients. Furthermore, while all patients treated in the HERACLES study and in the EAP were resistant to their previous lines of treatment, including ruxolitinib for patients treated in the EAP, we can’t exclude a delayed beneficial effect of these treatments, also contributing to the higher response rate in these patients. 重要的是，EAP包括9名患有SD-GI-aGvHD的患者，他们都在D28时获得了应答，这也是这些患者应答率较高的原因。此外，虽然所有接受 HERACLES 研究和 EAP 治疗的患者都对之前的治疗方案产生了耐药性，包括接受 EAP 治疗的患者对 Ruxolitinib 产生了耐药性，但我们不能排除这些治疗方案的延迟获益效应，这也是这些患者获得较高应答率的原因之一。
Similarly, the OS was higher in the EAP, being 49%49 \% at 6 months and 38%38 \% at 12 months, versus 28%28 \% at 6 months and 25%25 \% at 1 year in the HERACLES trial. The result may appear disappointing for the HERACLES trial, compared with the 49%49 \% weighted average 6 -month survival 同样，EAP的OS较高，6个月时为 49%49 \% ，12个月时为 38%38 \% ，而HERACLES试验的6个月时为 28%28 \% ，1年时为 25%25 \% 。与 49%49 \% 加权平均 6 个月存活率相比，HERACLES 试验的结果可能令人失望。
Table 3: Baseline demographics and clinical char patients within the compassionate use/expanded 表 3：恩恤用药/扩大用药范围内患者的基线人口统计学特征和临床症状
racteristics of treated access program. 治疗方案的特点。
Treated patients (n=52)
(Continued from previous column)
Calcineurin inhibitor 4 (8%)
MMF 2(4%)
mTOR inhibitor 2 (4%)
Anti-IL2R 2 (4%)
Vedolizumab 1 (2%)
Ibrutinib 1(2%)
Thalidomide 1(2%)
Steroid status
SR-aGvHD 43(83%)
SD-aGvHD 9 (17%)
Number of patients whose last line of treatment before MaaT013 was ruxolitinib 18 (%)
Median time between ruxolitinib initiation and first administration of MaaT013 in those patients, days
Median [range] 31 (16-613)
GvHD grading, n (%) )^(a,20)
I 0
II 3 (6%)
III 49 (94%)
IV 0
Stage skin GvHD
Stage 0 32 (68%)
Stage 1 8 (17%)
Stage 2 3 (6%)
Stage 3^(b) 4 (9%)
Stage 4 0
Missing data 5
Stage liver GvHD
Stage 0 41(87%)
Stage 1 4 (9%)
Stage 2 2 (4%)
Stage 3 0
Stage 4 0
Missing data 5
Stage gut GvHD
Stage 0 0
Stage 1 4 (8%)
Stage 2 12 (23%)
Stage 3 17 (33%)
Stage 4 19 (37%)
alloHCT: allogeneic haematopoietic cell transplantation; ATG: Anti-thymocyte globulin; PTCy: Post-transplant cyclophosphamide; CSA: Cyclosporin A; MMF: Mycophenolate mofetil; MTX: Methotrexate; CS: Corticosteroids; TNF, tumor necrosis factor; ECP, extracoporeal photopheresis; mTOR, mammalian target of rapamycin; IL2R, interleukin-2 receptor; SR: Steroid refractory; SD: Steroid dependance; aGvHD: acute graft-versus-host disease. ^("ar ") One patient had a dendritic cell acute leukaemia. ^("b ") One patient with overlap syndrome had severe skin chronic GvHD.
Table 3: Baseline demographics and clinical char patients within the compassionate use/expanded racteristics of treated access program.| | Treated patients $(n=52)$ |
| :---: | :---: |
| (Continued from previous column) | |
| Calcineurin inhibitor | 4 (8%) |
| MMF | $2(4 \%)$ |
| mTOR inhibitor | 2 (4%) |
| Anti-IL2R | 2 (4%) |
| Vedolizumab | 1 (2%) |
| Ibrutinib | $1(2 \%)$ |
| Thalidomide | $1(2 \%)$ |
| Steroid status | |
| SR-aGvHD | $43(83 \%)$ |
| SD-aGvHD | 9 (17%) |
| Number of patients whose last line of treatment before MaaT013 was ruxolitinib | 18 (%) |
| Median time between ruxolitinib initiation and first administration of MaaT013 in those patients, days | |
| Median [range] | 31 (16-613) |
| GvHD grading, n (%) $)^{a, 20}$ | |
| I | 0 |
| II | 3 (6%) |
| III | 49 (94%) |
| IV | 0 |
| Stage skin GvHD | |
| Stage 0 | 32 (68%) |
| Stage 1 | 8 (17%) |
| Stage 2 | 3 (6%) |
| Stage $3^{b}$ | 4 (9%) |
| Stage 4 | 0 |
| Missing data | 5 |
| Stage liver GvHD | |
| Stage 0 | $41(87 \%)$ |
| Stage 1 | 4 (9%) |
| Stage 2 | 2 (4%) |
| Stage 3 | 0 |
| Stage 4 | 0 |
| Missing data | 5 |
| Stage gut GvHD | |
| Stage 0 | 0 |
| Stage 1 | 4 (8%) |
| Stage 2 | 12 (23%) |
| Stage 3 | 17 (33%) |
| Stage 4 | 19 (37%) |
| alloHCT: allogeneic haematopoietic cell transplantation; ATG: Anti-thymocyte globulin; PTCy: Post-transplant cyclophosphamide; CSA: Cyclosporin A; MMF: Mycophenolate mofetil; MTX: Methotrexate; CS: Corticosteroids; TNF, tumor necrosis factor; ECP, extracoporeal photopheresis; mTOR, mammalian target of rapamycin; IL2R, interleukin-2 receptor; SR: Steroid refractory; SD: Steroid dependance; aGvHD: acute graft-versus-host disease. ${ }^{\text {ar }}$ One patient had a dendritic cell acute leukaemia. ${ }^{\text {b }}$ One patient with overlap syndrome had severe skin chronic GvHD. | |
| Table 3: Baseline demographics and clinical char patients within the compassionate use/expanded | racteristics of treated access program. |
must highlight that the HERACLES study and the EAP focus on GI-aGvHD which is a well-established risk factor for mortality. ^(28){ }^{28} Furthermore, in the HERACLES study 必须强调的是，HERACLES 研究和 EAP 重点关注的是消化道-aGvHD，这是一个公认的死亡风险因素。 ^(28){ }^{28} 此外，在HERACLES研究中
all but 3 patients have a high-risk Minnesota risk score ^(22){ }^{22} and all patients were high-risk based on the MAP score, associated with a 12-month OS ranging from 14 to 34%34 \% in early resistant patients. ^(23){ }^{23} Therefore, our study focusing on GI-aGvHD resistant to first line treatment with steroids, reported an OS within the range expected in this high-risk population. In fact, none of the deaths were related to MaaT013 and all causes of death, relapse of the underlying disease, GvHD or severe infection, represent the most common cause of death in this patient population. ^(29){ }^{29} 除3例患者外，所有患者的明尼苏达风险评分均为高风险 ^(22){ }^{22} ，根据MAP评分，所有患者均为高风险，这与早期耐药患者的12个月OS从14到 34%34 \% 不等有关。 ^(23){ }^{23} 因此，我们的研究侧重于对类固醇一线治疗耐药的消化道-aGvHD，报告的OS在这一高风险人群的预期范围内。事实上，所有死亡病例都与MaaT013无关，所有死亡原因、基础疾病复发、GvHD或严重感染都是该患者群体中最常见的死亡原因。 ^(29){ }^{29}
More specifically regarding safety of MaaT013, with a follow-up of SAE of one year, five infectious complications, including 3 sepsis, could not be excluded from being related to the study procedure in HERACLES, and in the EAP, 11 bacteraemia/sepsis were reported among 52 treated patients. For patients included in HERACLES, shotgun sequencing analyses of the identified strains suggest that none were found in MaaT013. Furthermore, our overall incidence of bacteraemia remains low, compared to an incidence of 31%31 \% in a bloodstream infection reported in patients with GI-aGvHD. ^(30){ }^{30} Furthermore, in a cohort of 127 patients with SR-aGvHD treated with a monoclonal antibody (inolimomab or etanercept), the cumulative incidence of bacteraemia was 63%63 \% at D100 and 73.5%73.5 \% at one year, with a one-year OS of 24%.^(31)24 \% .{ }^{31} Overall, this suggests that, on the contrary, MaaT013 may have a protective effect on bacterial translocation. Of note, we recently reported that autologous FMT after induction chemotherapy for acute myeloid leukaemia decreases the presence of antibiotic resistance genes carriage, and was associated with a low incidence of microbiologically and clinically defined infection during consolidation chemotherapy. ^(32){ }^{32} 更具体地说，关于 MaaT013 的安全性，在一年的 SAE 随访中，HERACLES 报告了 5 例感染性并发症（包括 3 例败血症），不能排除与研究过程有关；而在 EAP 中，52 名接受治疗的患者中报告了 11 例菌血症/败血症。对于纳入 HERACLES 的患者，对确定的菌株进行的射枪测序分析表明，在 MaaT013 中未发现任何菌株。此外，我们的菌血症总发生率仍然很低，而据报道，在消化道-aGvHD患者中，血流感染的发生率为 31%31 \% 。 ^(30){ }^{30} 此外，在一组接受单克隆抗体（依诺莫单抗或依那西普）治疗的127例SR-aGvHD患者中，菌血症的累积发生率在D100时为 63%63 \% ，一年时为 73.5%73.5 \% ，一年的OS为 24%.^(31)24 \% .{ }^{31} 总体而言，这表明，相反，MaaT013可能对细菌易位有保护作用。值得注意的是，我们最近报道，急性髓性白血病诱导化疗后的自体 FMT 可减少抗生素耐药基因的携带，并与巩固化疗期间微生物和临床定义的感染发生率低有关。 ^(32){ }^{32}
In responding patients at D28, we observed an increase in microbiota richness and diversity, particularly in butyrate-producing bacteria. Regarding engraftment, the percentage of bacteria originating from MaaT013 was higher in R compared to NR patients, nevertheless this increase did not reach statistical significance, possibly related to the small number of patients, therefore limiting the relevance of our study. Furthermore, the lower number of MaaT013 administered in NR patients may have contributed to the lower engraftment and to the detection of bacteria from fewer donors in the NR patients, further limiting our findings. Nevertheless, our results show that multiple donors engrafted in R patients, confirming that pooling of donors is feasible and results in engraftment of multiple donors. While we can’t exclude that we could have achieved similar outcomes with use of a single high diversity donor, identification of a sufficient number of high diversity donors may be difficult to achieve for use in routine practice. Finally, we found that RR and NR patients exhibited different microbiota and inflammatory patterns over 在有反应的患者中，我们观察到微生物群的丰富性和多样性有所增加，尤其是产丁酸的细菌。在移植方面，与 NR 患者相比，R 患者中来自 MaaT013 的细菌比例较高，但这一增长未达到统计学意义，这可能与患者人数较少有关，因此限制了我们研究的相关性。此外，NR 患者接受的 MaaT013 数量较少，这可能是导致 NR 患者移植率较低和检测到来自较少供体的细菌的原因，从而进一步限制了我们的研究结果。尽管如此，我们的研究结果表明，多个供体在 R 型患者中发生了接种，这证实了集中供体是可行的，而且会导致多个供体发生接种。虽然我们不能排除使用单个高分辨率供体也能获得类似结果的可能性，但在常规实践中可能难以确定足够数量的高分辨率供体。最后，我们发现 RR 和NR患者的微生物群和炎症模式不同于 RR 。
time. In fact, in R patients, we observed increased levels of beneficial bacteria, in particular butyrateproducers, along with an increased level of SCFAs and secondary bile acids. In contrast, NR patients exhibited increased levels of pathogenic proinflammatory bacteria along with increased systemic inflammatory parameters. While we can’t exclude that similar results would be achieved with other GvHD treatments, these findings suggest a systemic effect of MaaT013 in R patients. Unfortunately, regarding the EAP, no stool sample was planned in those patients. Therefore, we were unable to confirm HERACLES findings in this cohort and to assess whether the higher response rate in the EAP was related to a better MaaT013 engraftment. 时间。事实上，在 R 型患者中，我们观察到有益细菌的数量有所增加，尤其是丁酸生产者，同时 SCFAs 和次级胆汁酸的数量也有所增加。相比之下，NR 患者的致病性促炎症细菌水平升高，全身炎症指标升高。虽然我们不能排除其他 GvHD 治疗方法也会产生类似的结果，但这些研究结果表明，MaaT013 对 R 型患者有全身性影响。遗憾的是，关于 EAP，我们没有计划对这些患者进行粪便采样。因此，我们无法证实 HERACLES 在该组患者中的发现，也无法评估 EAP 中较高的反应率是否与 MaaT013 较好的接种有关。
This was the largest study performed so far in this setting, with 76 patients receiving pooled faecal microbiota MaaT013 for treatment of SR or SD-GIaGvHD. While we acknowledge that a randomised control trial would be the gold standard to assess MaaT013 efficacy, our results suggest that this approach is safe and may be effective for very high-risk immunocompromised patients. Interestingly, clinical response in SR-GI-aGvHD was linked to the increased microbiota richness resulting from the engraftment of multidonor material. A single-arm non-randomised phase III clinical trial is now ongoing in patients with SR or SD-aGvHD and resistance or intolerance to ruxolitinib (NCT04769895). 这是迄今为止在这种情况下开展的规模最大的研究，共有 76 名患者接受了集合粪便微生物群 MaaT013 治疗 SR 或 SD-GIaGvHD。虽然我们承认随机对照试验是评估 MaaT013 疗效的黄金标准，但我们的研究结果表明，这种方法是安全的，而且可能对极高风险的免疫功能低下患者有效。有趣的是，SR-GI-aGvHD 患者的临床反应与多原核材料移植后微生物群丰富度的增加有关。目前，一项针对SR或SD-aGvHD患者以及对鲁索利替尼耐药或不耐受患者的单臂非随机III期临床试验正在进行中（NCT04769895）。
Contributors 贡献者
FM and MM designed the study in collaboration with the legal sponsor MaaT Pharma. FM, ML, AH, TC, SG, FL, LM, CO, AC, MPK., DD, JBM, JC, CR, EH, ED, KB, MV, PC, HLW, CM MAC, CEB, VC, SC, PC, EH and MM included and treated patients. FM, BG, JD, EPr, CG, and EPI collected and analysed data. FM and MM interpreted data. FM wrote the first draft of the manuscript. All others authors provided editorial comments on the manuscript. All authors approved the manuscript. FM and EPl accessed and verified the underlying data. FM 和 MM 与法定赞助商 MaaT Pharma 合作设计了这项研究。FM、ML、AH、TC、SG、FL、LM、CO、AC、MPK.、DD、JBM、JC、CR、EH、ED、KB、MV、PC、HLW、CM MAC、CEB、VC、SC、PC、EH 和 MM 纳入并治疗患者。FM、BG、JD、EPr、CG 和 EPI 收集和分析数据。FM 和 MM 解释数据。FM 撰写了手稿初稿。所有其他作者对手稿提出了编辑意见。所有作者批准了手稿。FM和EPl获取并验证了基础数据。
Data sharing statement 数据共享声明
Sequence data that support the findings have been deposited in the NCBI SRA database under the BioProject accession PRJNA801217, and available at https://www.ncbi.nlm.nih.gov/bioproject/801217. All the other data supporting the findings of this study are available within the article and its Supplementary Information files and from the corresponding author upon reasonable request. Source data are provided with this paper. 支持研究结果的序列数据已存入 NCBI SRA 数据库，生物项目登录号为 PRJNA801217，可在 https://www.ncbi.nlm.nih.gov/bioproject/801217 上查阅。支持本研究结果的所有其他数据可在文章及其补充信息文件中获取，也可向相应作者索取。本文随附原始数据。
Declaration of interests 利益声明
FM reports honoraria from Therakos/Mallinckrodt, Sanofi and Novartis. TC reports consulting fee from Celgene, Abbvie, Jazz Pharma, Roche, Novartis, Agios and Servier and honoraria from Novartis, Amgen, Astellas, Celgene/BMS and Gilead/Kite. ED report payment for expert testimony from Astellas and support for attending meetings and/or travel from Neovii and Novartis. MJGTV report grants or contracts from MSD, Heel, Biontech and Roche, consulting fees from MaaT Pharma, Tillots, MSD/Merck, Roche and EUMEDICA and payment or honoraria from Merck/MSD, 3M, Ferring, Astellas, Uniklinik Karlsruhe, Uniklinik Köln, Akademie für Infektionsmedizin, Klinikum Essen, Pfizer, Universitätsklinikum Heidelberg, Uniklinik Frankfurt, Landesärztekammer Hessen, Janssen, Intitur Merieux, Forum für medizinische Fortbildung Gmbh, Universitätsklinikum Freiburg, Berliner FM 报告了 Therakos/Mallinckrodt、赛诺菲和诺华的酬金。TC 报告了 Celgene、Abbvie、Jazz Pharma、Roche、Novartis、Agios 和 Servier 的咨询费，以及 Novartis、Amgen、Astellas、Celgene/BMS 和 Gilead/Kite 的酬金。ED 报告称，Astellas 支付了专家证词费用，Neovii 和 Novartis 支付了出席会议和/或差旅费用。MJGTV 报告了 MSD、Heel、Biontech 和罗氏的资助或合同，MaaT Pharma、Tillots、MSD/Merck、罗氏和 EUMEDICA 的咨询费，以及 Merck/MSD、3M、Ferring、Astellas、Uniklinik Karlsruhe、Uniklinik Köln、Akademie für Infektionsmedizin、Klinikum Essen、Pfizer、Universitätsklinikum Heidelberg、Uniklinikfrankfurt、Landesärztekammer Hessen、Janssen、Intitur Merieux、Forum für medizinische Fortbildung Gmbh、Universitätsklinikum Freiburg、Berliner
Dialy Seminar, ADKA, Falk Foundation, St. Johannes Hospital, DiaLog Service, CED Service, Ärztekammer Niedersachsen, St. Josef Hospital, Limbach Gruppe SE, SUMIT OXFORD Ltd. EUMEDICA Kit Kongress, Tillots Pharma, Helios Kliniken, Lahn-Dill-Kliniken, GILEAD and Klinikum Leverkusen. CEB report honoraria from Astellas. EH report grants from Medac, payment for expert testimony from Novartis, participation on Advisory Board of Maat Pharma and Pharmabiome. JD is cofounder and member of the advisory board of MaaT Pharma, report consulting fees from MaaT Pharma, support for attending meetings and/or travel from MaaT Pharma and shares from MaaT Pharma. EPr, CG and EPl are MaaT Pharma employees and have patents as part of company inventor compensation policy MM reports grants and consulting fees from Sanofi, Novartis, and Janssen, honoraria from Sanofi, Novartis, Janssen, and MaaT Pharma, participation on Advisory Board of Janssen and leadership role in The European Group for Blood & Marrow Transplantation. All other authors declare no competing interests. Dialy Seminar、ADKA、Falk Foundation、St. Johannes Hospital、DiaLog Service、CED Service、Ärztekammer Niedersachsen、St. Josef Hospital、Limbach Gruppe SE、SUMIT OXFORD Ltd、EUMEDICA Kit Kongress、Tillots Pharma、Helios Kliniken、Lahn-Dill-Kliniken、GILEAD 和 Klinikum Leverkusen。EUMEDICA Kit Kongress、Tillots Pharma、Helios Kliniken、Lahn-Dill-Kliniken、GILEAD 和 Klinikum Leverkusen。CEB 报告了安斯泰来公司的酬金。EH 报告获得了 Medac 公司的资助、诺华公司的专家证言酬金，并加入了 Maat Pharma 和 Pharmabiome 公司的顾问委员会。JD 是 MaaT Pharma 公司的共同创始人和顾问委员会成员，报告了 MaaT Pharma 公司的咨询费、MaaT Pharma 公司对其出席会议和/或出差的资助以及 MaaT Pharma 公司的股份。EPr、CG和EPl是MaaT Pharma公司的员工，拥有公司发明者补偿政策规定的专利。MM报告了赛诺菲、诺华和杨森公司提供的资助和咨询费，赛诺菲、诺华、杨森和MaaT Pharma公司提供的酬金，在杨森公司顾问委员会中的参与情况，以及在欧洲血液与骨髓移植组织中的领导角色。所有其他作者均未声明任何利益冲突。
Acknowledgements 致谢
We thank the “Tumorothèque” of the Clinical Haematology Department at Saint-Antoine hospital for logistical support for handling of samples. This study was presented in part during the 63rd ASH annual meeting held in Atlanta, GA, USA, Dec 11-14, 2021. 感谢圣安东尼医院临床血液学部 "Tumorothèque "为样本处理提供的后勤支持。本研究的部分内容已于 2021 年 12 月 11-14 日在美国佐治亚州亚特兰大市举行的第 63 届 ASH 年会上发表。
Appendix A. Supplementary data 附录 A.补充数据
Supplementary data related to this article can be found at https://doi. org/10.1016/j.eclinm.2023.102111. 与本文相关的补充数据见https://doi。org/10.1016/j.eclinm.2023.102111。
References 参考资料
1 Yu J, Parasuraman S, Shah A, Weisdorf D. Mortality, length of stay and costs associated with acute graft-versus-host disease during hospitalization for allogeneic hematopoietic stem cell transplantation. Curr Med Res Opin. 2019;35(6):983-988. 1 Yu J、Parasuraman S、Shah A、Weisdorf D.异基因造血干细胞移植住院期间急性移植物抗宿主疾病相关的死亡率、住院时间和费用。Curr Med Res Opin.2019;35(6):983-988.
2 Farhadfar N, Leather HL, Wang S, et al. Severity of acute graft-versushost disease and associated healthcare resource utilization, cost, and outcomes. Transplant Cell Ther. 2021;27(12):1007.e1-1007.e8. 2 Farhadfar N, Leather HL, Wang S, et al. 急性移植物抗宿主疾病的严重程度与相关的医疗资源利用率、成本和结果。Transplant Cell Ther.2021;27(12):1007.e1-1007.e8.
3 Martin PJ, Rizzo JD, Wingard JR, et al. First- and second-line systemic treatment of acute graft-versus-host disease: recommendations of the American Society of Blood and Marrow Transplantation. Biol Blood Marrow Transplant. 2012;18(8):1150-1163. 3 Martin PJ、Rizzo JD、Wingard JR 等：《急性移植物抗宿主病的一线和二线系统治疗：美国血液和骨髓移植学会的建议》。Biol Blood Marrow Transplant.2012;18(8):1150-1163.
4 Van Lint MT, Uderzo C, Locasciulli A, et al. Early treatment of acute graft-versus-host disease with high- or low-dose 6-methylprednisolone: a multicenter randomized trial from the Italian Group for Bone Marrow Transplantation. Blood. 1998;92(7):2288-2293. 4 Van Lint MT、Uderzo C、Locasciulli A 等：用高剂量或低剂量 6-甲基强的松龙早期治疗急性移植物抗宿主病：意大利骨髓移植小组的一项多中心随机试验。血液。1998;92(7):2288-2293.
5 Alousi AM, Weisdorf DJ, Logan BR, et al. Etanercept, mycophenolate, denileukin, or pentostatin plus corticosteroids for acute graft-versus-host disease: a randomized phase 2 trial from the Blood and Marrow Transplant Clinical Trials Network. Blood. 2009;114(3):511-517. 5 Alousi AM, Weisdorf DJ, Logan BR, et al. Etanercept, mycophenolate, denileukin, or pentostatin plus corticosteroids for acute graft-versus-host disease: a randomized phase 2 trial from the Blood and Marrow Transplant Clinical Trials Network.血液。2009;114(3):511-517.
6 Bolanos-Meade J, Logan BR, Alousi AM, et al. Phase 3 clinical trial of steroids/mycophenolate mofetil vs steroids/placebo as therapy for acute GVHD: BMT CTN 0802. Blood. 2014;124(22):3221-3227. quiz 335 . 6 Bolanos-Meade J、Logan BR、Alousi AM 等：类固醇/霉酚酸酯与类固醇/安慰剂治疗急性 GVHD 的 3 期临床试验：BMT CTN 0802。Blood.2014；124（22）：3221-3227 。
7 Weisdorf D, Haake R, Blazar B, et al. Treatment of moderate/severe acute graft-versus-host disease after allogeneic bone marrow transplantation: an analysis of clinical risk features and outcome. Blood. 1990;75(4):1024-1030. 7 Weisdorf D, Haake R, Blazar B, et al. 异体骨髓移植后中度/重度急性移植物抗宿主疾病的治疗：临床风险特征和结果分析。血液。1990;75(4):1024-1030.
8 Taur Y, Jenq RR, Perales MA, et al. The effects of intestinal tract bacterial diversity on mortality following allogeneic hematopoietic stem cell transplantation. Blood. 2014;124(7):1174-1182. 8 Taur Y、Jenq RR、Perales MA 等：肠道细菌多样性对异体造血干细胞移植后死亡率的影响。Blood.2014;124(7):1174-1182.
9 Peled JU, Gomes ALC, Devlin SM, et al. Microbiota as predictor of mortality in allogeneic hematopoietic-cell transplantation. NN Engl JJ Med. 2020;382(9):822-834. 9 Peled JU, Gomes ALC, Devlin SM, et al. 异体造血细胞移植中预测死亡率的微生物群。 NN Engl JJ Med.2020;382(9):822-834.
10 Shono Y, van den Brink MRM. Gut microbiota injury in allogeneic haematopoietic stem cell transplantation. Nat Rev Cancer. 2018;18(5):283-295. 10 Shono Y，van den Brink MRM。异基因造血干细胞移植中的肠道微生物群损伤。Nat Rev Cancer.2018;18(5):283-295.
11 Shono Y, Docampo MD, Peled JU, et al. Increased GVHD-related mortality with broad-spectrum antibiotic use after allogeneic 11 Shono Y, Docampo MD, Peled JU, et al.
hematopoietic stem cell transplantation in human patients and mice. Sci Transl Med. 2016;8(339):339ra71. 在人类患者和小鼠中进行造血干细胞移植。Sci Transl Med.2016;8(339):339ra71.
12 Stein-Thoeringer CK, Nichols KB, Lazrak A, et al. Lactose drives Enterococcus expansion to promote graft-versus-host disease. Science. 2019;366(6469):1143-1149. 12 Stein-Thoeringer CK、Nichols KB、Lazrak A 等人。乳糖驱动肠球菌扩增，促进移植物抗宿主疾病。科学。2019;366(6469):1143-1149.
13 Jenq RR, Taur Y, Devlin SM, et al. Intestinal Blautia is associated with reduced death from graft-versus-host disease. Biol Blood Marrow Transplant. 2015;21(8):1373-1383. 13 Jenq RR、Taur Y、Devlin SM 等人.肠布劳狄与减少移植物抗宿主病死亡有关。生物血液骨髓移植。2015;21(8):1373-1383.
14 Mathewson ND, Jenq R, Mathew AV, et al. Gut microbiomederived metabolites modulate intestinal epithelial cell damage and mitigate graft-versus-host disease. Nat Immunol. 2016;17(5):505-513. 14 Mathewson ND、Jenq R、Mathew AV 等人. 肠道微生物衍生的代谢物可调节肠道上皮细胞损伤并减轻移植物抗宿主疾病。Nat Immunol.2016;17(5):505-513.
15 van Nood E, Vrieze A, Nieuwdorp M, et al. Duodenal infusion of donor feces for recurrent Clostridium difficile. NN Engl J Med. 2013;368(5):407-415. 15 van Nood E, Vrieze A, Nieuwdorp M, et al.十二指肠输注供体粪便治疗复发性艰难梭菌。 NN 英国医学杂志。2013;368(5):407-415.
16 Kakihana K, Fujioka Y, Suda W, et al. Fecal microbiota transplantation for patients with steroid-resistant acute graft-versus-host disease of the gut. Blood. 2016;128(16):2083-2088. 16 Kakihana K、Fujioka Y、Suda W 等.粪便微生物群移植治疗耐类固醇急性肠道移植物抗宿主疾病患者.Blood.2016;128(16):2083-2088.
17 Qi X, Li X, Zhao Y, et al. Treating steroid refractory intestinal acute graft-vs.-host disease with fecal microbiota transplantation: a pilot study. Front Immunol. 2018;9:2195. 17 Qi X、Li X、Zhao Y 等：粪便微生物群移植治疗类固醇难治性肠道急性移植物抗宿主病：一项试验研究。Front Immunol.2018;9:2195.
18 van Lier YF, Davids M, Haverkate NJE, et al. Donor fecal microbiota transplantation ameliorates intestinal graft-versus-host disease in allogeneic hematopoietic cell transplant recipients. Sci Transl Med. 2020;12(556):eaaz8926. 18 van Lier YF、Davids M、Haverkate NJE 等：捐赠者粪便微生物群移植可改善异体造血细胞移植受者的肠道移植物抗宿主疾病。Sci Transl Med.2020;12(556):eaaz8926.
19 ANSM Agence nationale de sécurité du médicament et des produits de santé. La transplantation de microbiote fécal et son encadrement dans les essais cliniques. 2016. 19 ANSM Agence nationale de sécurité du médicament et des produits de santé.La transplantation de microbiote fécal et son encadrement dans les essais cliniques.2016.
20 Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus conference on acute GVHD grading. Bone Marrow Transplant. 1995;15(6):825-828. 20 Przepiorka D、Weisdorf D、Martin P 等人。1994 年急性 GVHD 分级共识会议。骨髓移植。1995;15(6):825-828.
21 Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Roy Stat Soc B. 1995;57(1):289-300. 21 Benjamini Y, Hochberg Y.控制误发现率：一种实用而强大的多重检验方法。J Roy Stat Soc B. 1995;57(1):289-300.
22 MacMillan ML, Robin M, Harris AC, et al. A refined risk score for acute graft-versus-host disease that predicts response to initial therapy, survival, and transplant-related mortality. Biol Blood Marrow Transplant. 2015;21(4):761-767. 22 MacMillan ML、Robin M、Harris AC 等.可预测对初始治疗的反应、存活率和移植相关死亡率的急性移植物抗宿主病的改良风险评分。生物血液骨髓移植。2015;21(4):761-767.
23 Major-Monfried H, Renteria AS, Pawarode A, et al. MAGIC biomarkers predict long-term outcomes for steroid-resistant acute GVHD. Blood. 2018;131(25):2846-2855. 23 Major-Monfried H、Renteria AS、Pawarode A 等：MAGIC 生物标志物可预测类固醇耐药急性 GVHD 的长期预后。Blood.2018;131(25):2846-2855.
24 Weber D, Oefner PJ, Hiergeist A, et al. Low urinary indoxyl sulfate levels early after transplantation reflect a disrupted microbiome and are associated with poor outcome. Blood. 2015;126(14): 1723-17281723-1728. 24 Weber D、Oefner PJ、Hiergeist A 等人.移植后早期尿液中吲哚硫酸酯水平低反映了微生物群紊乱并与不良预后相关。Blood.2015;126(14): 1723-17281723-1728 。
25 Mohty M, Holler E, Jagasia M, et al. Refractory acute graft-versushost disease: a new working definition beyond corticosteroid refractoriness. Blood. 2020;136(17):1903-1906. 25 Mohty M, Holler E, Jagasia M, et al. 难治性急性移植物抗宿主病：超越皮质类固醇难治性的新工作定义。Blood.2020;136(17):1903-1906.
26 Jagasia M, Perales MA, Schroeder MA, et al. Ruxolitinib for the treatment of steroid-refractory acute GVHD (REACH1): a multicenter, open-label, phase 2 trial. Blood. 2020;135(20):1739-1749. 26 Jagasia M, Perales MA, Schroeder MA, et al. Ruxolitinib 用于治疗类固醇难治性急性 GVHD (REACH1)：一项多中心、开放标签、2 期试验。Blood.2020;135(20):1739-1749.
27 Zeiser R, von Bubnoff N, Butler J, et al. Ruxolitinib for glucocorticoid-refractory acute graft-versus-host disease. N Engl J Med. 2020;382(19):1800-1810. 27 Zeiser R, von Bubnoff N, Butler J, et al. Ruxolitinib 治疗糖皮质激素难治性急性移植物抗宿主病。N Engl J Med.2020;382(19):1800-1810.
28 Castilla-Llorente C, Martin PJ, McDonald GB, et al. Prognostic factors and outcomes of severe gastrointestinal GVHD after allogeneic hematopoietic cell transplantation. Bone Marrow Transplant. 2014;49(7):966-971. 28 Castilla-Llorente C、Martin PJ、McDonald GB 等. 异基因造血细胞移植后严重胃肠道 GVHD 的预后因素和结果。骨髓移植。2014;49(7):966-971.
29 Khoury HJ, Wang T, Hemmer MT, et al. Improved survival after acute graft-versus-host disease diagnosis in the modern era. Haematologica. 2017;102(5):958-966. 29 Khoury HJ、Wang T、Hemmer MT 等人.现代急性移植物抗宿主疾病诊断后生存率的提高.血液病学杂志（Haematologica）.2009 年.Haematologica.2017;102(5):958-966.
30 Mori Y, Yoshimoto G, Nishida R, et al. Gastrointestinal graft-versus-host disease is a risk factor for postengraftment bloodstream infection in allogeneic hematopoietic stem cell transplant recipients. Biol Blood Marrow Transplant. 2018;24(11):2302-2309. 30 Mori Y, Yoshimoto G, Nishida R, et al. 胃肠道移植物抗宿主疾病是异体造血干细胞移植受者移植后血流感染的危险因素。Biol Blood Marrow Transplant.2018;24(11):2302-2309.
31 García-Cadenas I, Rivera I, Martino R, et al. Patterns of infection and infection-related mortality in patients with steroid-refractory acute graft versus host disease. Bone Marrow Transplant. 2017;52(1):107-113. 31 García-Cadenas I、Rivera I、Martino R 等：类固醇难治性急性移植物抗宿主疾病患者的感染模式和感染相关死亡率。骨髓移植。2017;52(1):107-113.
32 Malard F, Vekhoff A, Lapusan S, et al. Gut microbiota diversity after autologous fecal microbiota transfer in acute myeloid leukemia patients. Nat Commun. 2021;12(1):3084. 32 Malard F、Vekhoff A、Lapusan S 等：急性髓性白血病患者自体粪便微生物群转移后的肠道微生物群多样性。Nat Commun.2021;12(1):3084.
*Corresponding author. Sorbonne Université, AP-HP, Centre de Recherche Saint-Antoine INSERM UMRs938, Service D’Hématologie Clinique et de Thérapie Cellulaire, Hôpital Saint Antoine, AP-HP, Paris, France. E-mail address: florent.malard@inserm.fr (F. Malard). *通讯作者。Sorbonne Université, AP-HP, Centre de Recherche Saint-Antoine INSERM UMRs938, Service D'Hématologie Clinique et de Thérapie Cellulaire, Hôpital Saint Antoine, AP-HP, Paris, France.电子邮件地址florent.malard@inserm.fr (F. Malard)。
@Florent_Malard (F. Malard). @Florent_Malard (F. Malard)。
improvement of 1 stage in the severity of GI-aGvHD\mathrm{GI}-\mathrm{aGvHD}. Abbreviations: aGvHD is for acute graft-versus-host disease; SR-aGvHD, steroid-resistant acute graft-versus-host disease; D, day; V, visit; M, month; PR, partial response; VGPR, very good partial response and CR, complete response). Dose stands for MaaT013 administration. Faeces and blood collection (peripheral blood mononuclear cells and plasma) was performed at D0, D2, D9, D16 and D28, before MaaT013 administration when sample collection was performed the day of MaaT013 administration. GI-aGvHD\mathrm{GI}-\mathrm{aGvHD} 的严重程度改善一个阶段。缩写：aGvHD 代表急性移植物抗宿主病；SR-aGvHD 代表类固醇耐药急性移植物抗宿主病；D 代表日；V 代表访问；M 代表月；PR 代表部分应答；VGPR 代表非常好的部分应答；CR 代表完全应答）。剂量代表 MaaT013 的给药剂量。粪便和血液采集（外周血单核细胞和血浆）在D0、D2、D9、D16和D28进行，在MaaT013用药前采集样本的时间为MaaT013用药当天。
Curtis similarity index: comparison between visits and MaaT013 based on the Bray-Curtis similarity at each visit (median and interquartile range are provided). Bray-Curtis similarity index calculated as 1 minus the Bray-Curtis dissimilarity index. (D) Engraftment measured at OTU level (median and interquartile range are provided). (E) Richness versus number of donors from whom specific OTUs engrafted (out of 3-6 donors). Clinical Gl response was evaluated at each visit. Ellipse size is set to include 70%70 \% of the number of observations for each group. CR, complete response; VGPR, very good partial response; PR, partial response. Response was evaluated at D28. To measure engraftment for each patient sample at D9, D16, and D28, we determined among the OTUs detected in the sample, what proportion of OTUs was found among the list of OTUs detected specifically in the product (1st panel), specifically in the patient at DO (second panel), in both the patient and the product (3rd panel), or in neither of them (values not shown). Statistical significance was evaluated using unpaired two-sided Wilcoxon tests performed for each parameter. The p-value is considered significant after False Discovery Rate (FDR) controlled multiple tests adjustment when it is lower than 0.05 (the p-values displayed on the figures are adjusted). 柯蒂斯相似性指数：根据每次就诊的布雷-柯蒂斯相似性（提供中位数和四分位数间距），比较就诊与 MaaT013 之间的差异。布雷-柯蒂斯相似性指数以 1 减布雷-柯蒂斯不相似性指数计算。(D) 在 OTU 水平测量的接种率（提供中位数和四分位数间距）。(E）丰富度与移植特定 OTU 的供体数量的关系（3-6 位供体）。每次就诊均评估临床 Gl 反应。椭圆大小设置为包括每组观察数的 70%70 \% 。CR，完全应答；VGPR，非常好的部分应答；PR，部分应答。反应在第 28 天进行评估。为了衡量每个患者样本在 D9、D16 和 D28 时的移植情况，我们确定了在样本中检测到的 OTUs，在产品（第一组）、DO 患者（第二组）、患者和产品（第三组）或两者中均未检测到的 OTUs 列表中所占的比例（数值未显示）。统计意义是通过对每个参数进行非配对双侧 Wilcoxon 检验来评估的。经过假发现率（FDR）控制的多重检验调整后，当 p 值低于 0.05 时，则认为 p 值具有显著性（图中显示的 p 值经过调整）。
© Overall survival in HERACLES. CR was defined as complete resolution of Gl-aGvHD manifestations. VGPR was defined as improvement of at least 2 stages in the severity of 
