This was a prospective, single-arm, open label, phase 2 study evaluating UCB infusion as an adjuvant consolidation regimen in elderly AML patients (ChiCTR-OPC-15006492). This trial was approved by the Human Ethics Committee of Ruijin Hospital affiliated to Shanghai Jiao Tong University School of Medicine (RJ-201646). The protocol was in accordance with Declaration of Helsinki Principles and International Conference on Harmonization (ICH) Good Clinical Practice (GCP) Guidelines. After the institutional review board approval, the trial recruited patients at Ruijin Hospital. Written informed consents were obtained from all enrolled patients.

The complete inclusion criteria are shown below: age ≥60 years old; newly diagnosed with de novo AML, examined by bone marrow according to the WHO 2022 criteria; ECOG performance score 0 to 2; echocardiography examination of Left Ventricular Ejection Fraction (LVEF) ≥ 50%; estimated glomerular filtration rate (eGFR) ≥ 60 mL/min, measured using the CKD-EPI formula; alanine aminotransferase (ALT) ≤ 2.5×upper limit of normal (ULN), aspartate aminotransferase (AST) ≤ 2.5×ULN and total bilirubin ≤ 1.5×ULN; documented CR after one or two cycles of induction therapies defined according to standard criteria; inability or unwillingness to undergo allo-HSCT; signed informed consent before admission to the study.

The complete exclusion criteria are shown below: acute promyelocytic leukemia (APL); secondary AML (sAML); subjects who had been pretreated with other investigational drugs and/or currently participate in any other clinical trials; subjects with known involvement of central nervous system (CNS); subjects with known history of human immunodeficiency virus (HIV) infection; subjects with known history of hepatitis B virus (HBV) or hepatitis C virus (HCV) infection; New York Heart Association (NYHA) functional classification higher than grade 2; subjects with chronic respiratory disease requiring continuous oxygen inhalation; subjects with other malignant tumors or hematological system diseases; subjects with uncontrolled systemic infection (viral, bacterial or fungal); subjects with known or suspected autoimmune diseases; subjects with known history of intolerance or allergy to congeneric drugs; inability or unwillingness to follow the required protocol procedures; familial, psychological, geographical or sociological factors potentially impeding compliance with the protocol procedures and follow-up schedules; any uncontrolled or serious medical disorders that, at the discretion of the investigators, may increase the risks related to study participants or drug administrations, impair the abilities of the patients to undergo protocol therapies or obstruct the data interpretation of the study.

The first patient was enrolled on 12 January 2015 and the last patient on 12 February 2022. Written informed consents were received from all participants or their legal representatives, before any study related tests or procedures. There were no protocol amendments during the conduct of the study. Data were collected by the investigators and analyzed by the authors. The authors are committed to the fidelity of the trial, and the accuracy and integrity of the data in the study protocol.

The post-remission treatments for elderly patients with AML consisted of two cycles of 15 mg/m2 low dose intravenous decitabine over 4 h per day for consecutive 5 days (day 1–5), 1.0 g/m2 intermediate-dose cytarabine at q12h for consecutive 2 days (day 6–7), combined with one unit UCB infusion on day 9 as an adjuvant consolidation therapy. High resolution HLA typing of HLA-A, HLA-B and HLA-DR loci was carried out for all enrolled participants. The UCB units were gained from Shandong Province Cord Blood Bank (SINOCORD), if they were serologically matched of at least 4 HLAs and contained more than 3 × 107 nucleated cells per kilogram of patient’s body weight before freezing.

No immunosuppression was given as prophylaxis for GVHD, unless aGVHD was documented or diagnosed clinically. Infection prophylaxis or other support therapies, for example G-CSF, were administered following the regular transplantation procedure. The hematopoietic cell transplantation comorbidity index (HCT-CI) was applied to evaluate the comorbidities prior to each cycle of consolidation therapy. The second cycle of consolidation therapy was repeated after bone marrow recovery. After two cycles of consolidation therapy, patients received demethylation treatment as maintenance therapy. No targeted drugs were used in this study due to the lack of availability.

Patients underwent medical history inquiry, physical examinations, and molecular profiling at enrollment. Laboratory tests, including coagulation parameters, complete blood count, urine protein and urinalysis, a full set of biochemistries, electrocardiogram (ECG), echocardiography, abdominal ultrasound scan, chest CT scan, karyotyping, cytological and MRD detection of bone marrow biopsy, were performed at baseline and periodically thereafter at the treating investigators’ discretion. Response was assessed after each cycle of consolidation. Follow-up assessments were carried out monthly for the first year, and every 3 months thereafter.

Leukemia associated immunophenotyping (LAIP) was determined at diagnosis by using different cell surface markers. Bone marrow cytology and MRD were assessed according to the previously described process by muti-parameter flow cytometry.46 The monoclonal antibodies against 20 antigens were as follow: CD2, cyCD3, CD4, CD7, CD11b, CD13, CD14, CD15, CD19, CD33, CD34, CD38, CD45, CD56, CD64, cyCD79a, CD117, HLA-DR, MPO, and TdT. MRD negativity was defined as MRD < 0.01%. The samples were considered as evaluable if they contained 100,000 cells or more. Otherwise, they were determined as unevaluable.

On the day 7 after UCB infusion, peripheral blood cells were collected from all patients and assessed for chimerism using standard cytogenetic and a semi-quantitative polymerase chain reaction based assay of short tandem repeats with the sensitivity of 1%.

The primary endpoint of this trial was OS, defined as the time interval from the date of diagnosis to the date of death from any cause, with censoring of patients known to be alive upon the last follow-up. Secondary endpoints included EFS, bone marrow MRD by flow cytometry, treatment-related AEs, and the times to platelet and neutrophil count recovery. EFS was determined as the time interval from the date of diagnosis to the date of occurrence of any following events, including relapse or death, whichever came first. MRD relapse was not considered as an event for EFS, as not all MRD positive patients end with relapse.9 NRM was defined as death during continuous CR1. Exploratory endpoint was blood-based biological characteristics exploratory study analyzed by scRNA-seq of matched pre- and post-UCB infusion samples.

All the patients enrolled were included in the safety analyses. Treatment-related AEs, including hematological and non-hematological AEs, were defined as those that occurred from the start of treatment. Early deaths were defined as deaths occurring within 30 days of treatment with this regimen. The severity of AEs was assessed following the Common Terminology Criteria for Adverse Events (CTCAE, v5.0) of National Cancer Institute during treatment. AEs monitoring began after the first patient enrollment and was monitored continuously until the last follow-up. Grade 1–2 were defined as mild events, while grade 3–4 were defined as severe events.

The cell suspension (cell viability >80%, measured by Count Star) was processed using the Chromium single-cell controller (10x Genomics) to produce single-cell gel beads in emulsion, based on the manufacturer’s instructions. The quality of complementary DNA (cDNA) was evaluated using an Agilent 4200 system (Agilent Technologies). The libraries were then prepared utilizing the Chromium Single Cell 5’ Reagent Kits (10x Genomics, v1.1), according to the manufacturer’s protocol. The single-cell libraries were run on a MGISEQ 2000 sequencer (MGI) at the sequencing platform of National Research Center for Translational Medicine at Shanghai.

Sequencing data were demultiplexed and converted to FASTQ format files using spliteBarcode (MGI, v2.0.0) software. The raw reads were then mapped to the GRCh38 reference genome using cellranger (10x Genomics, v5.0.1) toolkit. In particular, cellranger count was employed to process the gene expression library sequencing data and generate the gene-barcode matrix.

The Seurat (v4.0.2) package in R (v4.1.0) was used for subsequent analyses.47,48,49 For quality control, the cells with percentage of mitochondrial genes more than 50%, detected genes less than 200 or more than 6000, were excluded from the analysis. Filtered data were log normalized using a scaling factor of 10,000. The most 2000 variable features per sample were selected using the variance stabilizing transformation (vst) method. To account for the batch effects across different samples, we utilized the canonical correlation analysis (CCA) approach within the Seurat package for data integration. The batch corrected and scaled data were used for dimensionality reduction.

Principal component analysis (PCA) was carried out to reduce the CCA merged data to the top 50 principal components (PCs). The first 30 PCs were utilized for further analysis based on the ElbowPlot. Clustering were identified using the shared nearest neighbor (SNN) graph-based model in the function FindClusters with resolution = 0.8. The same PCs (top 30) were utilized to generate Uniform Manifold Approximation and Projection (UMAP)50 for data visualization. Clusters were annotated by the expression patterns of canonical marker genes.

To identify the differentially expressed markers for each cluster, we utilized Wilcoxon rank sum test embedded in the functions FindAllMarkers and FindMarkers for multiple and two condition comparisons, respectively. Significant differentially expressed markers were identified as those with log2 fold change of average expression larger than 0.25, adjusted P values by Bonferroni-Hochberg less than 0.05, and expression present in more than 10% of cells. Gene ontology (GO) enrichment analysis was carried out by utilizing clusterProfiler (v4.6.2).51

Single cells assigned to CD8+ T cell clusters from all samples were employed for pseudotime analysis and diffusion map. We applied Monocle2 (v2.22.0) for the trajectory analysis with using the parameters of percentage of cells expressing each feature larger than 0.1 and genes expressed in at least 100 cells.28 The DEGs for each cluster were determined based on a likelihood ratio test. The significantly DEGs identified as those with adjusted P values less than 0.01 were selected as the ordering genes for trajectory reconstruction using the DDRTree nonlinear reconstruction algorithm.

Clinicopathological characteristics of patients were summarized by utilizing frequencies (%) for categorical variables and medians (range) for continuous variables. Fisher’s exact test was used to compare the associations for categorical variables versus categorical variables. Wilcoxon rank sum test was employed to compare categorical variables versus continuous variables. False discovery rate (FDR) corrected by Benjamini-Hochberg approach was applied to adjust the p value for multiple testing unless otherwise specified. Kaplan–Meier survival curves were utilized to evaluate the probabilities of OS and EFS. Log-rank test was performed to compare the survival curves. Asterisks define significance levels. All statistical tests were two-sided unless otherwise specified. Statistical analyses were carried out using R (v4.1.0) packages.