Patient and transplant characteristics

Patients and transplant characteristics are summarized in Table 1. Among 306 patients, 112 (36.6%) received transplantation for AML and 194 (63.1%) for ALL. Median age at UCBT was 13.87 years (range, 1-64 years) for patients with AML and 11.89 years (range, 1-50 years) for those with ALL. In the AML group, 65 patients (58%) were transplanted in CR1, 57 (18.7%) in CR2, and 25 (22.3%) in advanced disease status. For the ALL group, 103 patients (53%) were transplanted in CR1, 57 (29.4%) in CR2, and 24 (12.4%) in advanced disease status. Overall, 248 patients (81%) had high-risk disease at diagnosis. Among all patients, 31 (10.1%) received UCB that was 6/6 HLA matched, 152 (49.7%) with a 5/6 HLA match, and 123 (40.2%) with a 4/6 HLA match. A total 118 patients (38.5%) had matched donor ABO compatibility, 101 (33%) had major ABO mismatch, and 87 (28.4%) had minor ABO mismatch. There were 203 patients (66.3% of the entire population, of which 69 had AML and 134 had ALL) who received a conditioning regimen of BUCY2 in combination with fludarabine. Another 103 patients (33.3% of the total population, of which 43 had AML and 60 with ALL) received a conditioning regimen of total body irradiation cyclophosphamide plus cytarabine. For patients with AML, the median total nucleated cell dose and CD34⁺ cell dose was 4.25 × 10⁷/kg (range: 1.71-17.27 × 10⁷/kg) and 2.18 × 10⁵/kg (range: 0.45-10.55 × 10⁵/kg), respectively. For ALL patients, the median total nucleated cell dose and CD34⁺ cell dose was 4.22 × 10⁷/kg (range: 1.69-13.5 × 10⁷/kg) and 2.25 × 10⁵/kg (range, 0.4-8.53 × 10⁵/kg), respectively.

Outcomes of transplantation

For the total population of patients with AML and ALL, the engraftment of neutrophil and platelet were similar. The 30-day cumulative incidences of neutrophil engraftment after UCBT in the AML and ALL groups were 95.5% (95% CI: 89.1-98.2) and 94.8% (95% CI: 90.5-97.2), respectively (P=0.718). The median time to neutrophil recovery was 17 (range, 12-30) days and 16 (range, 10-31) days, respectively (Table 2). Fourteen patients did not engraft after UCBT (AML: 5 and ALL: 9); four of those patients—all of whom had ALL-died from severe infection and cerebral hemorrhage (two patients each). The 120-day cumulative incidences of platelet engraftment for AML and ALL was 83.9% (95% CI: 75.6-89.7) and 82.5% (95%CI: 76.3-87.2; P=0.733) and the median time to platelet recovery was 36 (range, 17-121) and 36 (range, 11-196) days, respectively.

With respect to age groups (≤ 14 years, 15-39 years, ≥ 40 years) (Table 3), adolescents and young adults group had higher engraftment rates for AML than ALL. The 30-day cumulative incidence of neutrophil recovery was higher in patients with AML than in those with ALL (97.9%, 95% CI: 76.8-99.8, vs. 91.1%, 95% CI: 78.7-96.4, P=0.02). In multivariate analysis, CD34⁺ cell dose and conditioning regimen were independent factors influencing neutrophil recovery. The cumulative incidence of platelet recovery on day 120 was also higher in AML than in ALL (87.2%, 95% CI: 72.8-94.3, vs. 73.2%, 95% CI: 59.2-83.1; P=0.03). In multivariate analysis, CD34⁺ cell dose was the only independent factor influencing platelet recovery (Table 4). There was no difference between AML and ALL in the groups of children and old adults.

Incidences of acute and chronic GVHD

For the total population of patients with AML and ALL, the cumulative incidences of acute GVHD (grade II-IV and grade III-IV) at 100 days were similar (grades II-IV: AML, 17.1% (95% CI: 10.8-24.7) and ALL, 14.5% (95% CI: 10.8-24.7), P=0.613; grades III-IV: AML, 11% (95% CI: 5.9-17.4) and children, 8% (95% CI: 4.9-12.7), P=0.489). The 2-years cumulative incidences of cGVHD in patients with AML and ALL were also similar (23%, (95% CI: 15.5-31.4) and 17%, (95% CI: 12.1-22.6, P=0.377). With grouping by age, cumulative incidences of aGVHD and cGVHD were still similar between AML and ALL in children, adolescents and young adults, and older adults.

Relapse and non-relapse mortality, overall survival, and GRFS

Fifty-four patients had relapse after UCBT (11 with AML and 43 with ALL). The cumulative incidence for relapse at 2 years in patients with ALL was higher than with AML (24%, 95% CI: 17.9-30.6, vs. 10.1% 95% CI: 5.1-17.1, P=0.01) (Figure 1). The median time to relapse was 12.7 (range, 1.1-59.6) months among AML and 10 (range, 1.5-34.8) months among ALL. In the analysis of age-group, both cumulative incidence of relapse and median time to relapse were significantly different between AML and ALL in adolescents and young adults group (relapse: 25%, 95% CI: 1.4-37.1, vs. 9%, 95% CI: 5.3-15.8, P=0.02); median time: 20.4 (range, 18.6-59.6) months vs. 8.1 (range, 2.8-17.7) months). There were no differences between AML and ALL in both children group and old adult group. In multivariate analysis, cGVHD was an independent factor influencing relapse.

Sixty patients died without relapse, 23 with AML and 37 with ALL. The median time to non-relapse mortality in AML and ALL was 301 (range: 45-1 044) days and 121 (range: 3-1 392) days, respectively. The cumulative incidence of non-relapse mortality at 2 years was 21.4%(95% CI: 14.1-29.7) for AML and 18.1%(95% CI: 13.0-23.9) for ALL (P=0.738). Causes of death included pulmonary infection (n=30), hemorrhage (n=4), multiorgan failure (n=22), and other causes (n=4). There was no significant difference in non-relapse mortality between AML and ALL by age groups.

The median follow-up duration was 21.6 (range, 3-80.7) months. Overall survival at 2 years in AML was similar to that in ALL (72%, 95% CI: 62-79.4, vs. 66%, 95% CI: 59.5-70.3, P=0.263). There was also no difference in overall survival between AML and ALL by age groups. The GRFS at 2 years in AML and ALL was 63.1% (95% CI: 52.9-71.7) and 56.2% (95% CI: 48.7-63.0), respectively (P=0.192). GRFS also showed no difference between AML and ALL by age groups. 

Immune reconstitution

Considering the effect of age on the development of the immune system, the immune reconstitution was compared among age groups. In the comparison of outcomes between AML and ALL, we analyzed age groups ≤ 14 years and 15-39 years with respect to CD4⁺ T cells, CD8⁺ T cells, natural killer cells, and Treg cells at 1, 2, 3, 4,6, 12, and 18 months after CBT (Figure 2).

The results showed that CD4⁺, CD8⁺, and Treg cells recovered with time after CBT whereas natural killer cells recovered more quickly (within 1 month) in patients in both age groups. In the group ≤ 14 years, CD4⁺ T cell counts in AML were lower than in ALL at 2 months after CBT whereas Treg cell counts in AML were higher than in ALL at 4 months post CBT. There was no difference seen in CD8⁺ and natural killer cells between AML and ALL. The proportion of CD4/CD8 T cells was lowest at 2 months and recovered at 4 months after CBT. 

In the age group 15-39 years, CD8⁺ T cell counts in AML at 4 months post CBT were higher than in ALL whereas CD4⁺, Treg, and natural killer cells showed no difference between AML and ALL. The proportion of CD4⁺/CD8⁺ T cells was lowest at 3 months and recovered at 12 months post CBT.

For choosing the better style of transplantation, we were interested in determining whether AML and ALL have similar outcomes among different age groups after single-unit CBT using the same conditioning regimen, such as myeloablative conditioning without antithymocyte globulin. As the biological differences of acute leukemia between adolescents and young adults (aged 15-39 years) and adults has been confirmed, more and more studies have focused on the outcomes of adolescents and young adults, in whom treatment strategies have remained controversial until now^[10]. Therefore, we compared the outcomes of AML and ALL after single-unit CBT using myeloablative conditioning without antithymocyte globulin among different groups based on age, in children, adolescents and young adults, and adults. Different to the findings of previous research, our results add new data regarding immune reconstitution.

中国组织工程研究杂志出版内容重点：干细胞；骨髓干细胞；造血干细胞；脂肪干细胞；肿瘤干细胞；胚胎干细胞；脐带脐血干细胞；干细胞诱导；干细胞分化；组织工程

Design 

Retrospective study.

Time and setting

This work was conducted at First Affiliated Hospital of University of Science and Technology of China (Anhui Provincial Hospital) from November 2011 to July 2018.

Participants

This retrospective registry-based study included 112 patients with AML and 194 with ALL in First Affiliated Hospital of University of Science and Technology of China (Anhui Provincial Hospital), between November 2011 and July 2018. Considering the difference in the age at diagnosis between AML and ALL, we used grouping by age to minimize the impact of this difference. The age groups were as follows: ≤ 14, 15-39, and ≥ 40 years.

All patients were treated according to our protocol, which is registered at www.chictr.org.cn (# ChiCTR-ONRC- 11001430). Written informed consent was obtained from all patients or their guardians prior to the start of conditioning. The transplant protocol was approved by the Ethics Committee of Anhui Provincial Hospital (approval No. 2011-089) on March 31, 2011.

Inclusion criteria

(1) All patients were diagnosed as having ALL or AML. (2) All patients had a suitable cord blood for transplantation. (3) All patients agreed to receive CBT.

Exclusion criteria

There were (1) impaired heart and renal function; (2) uncontrolled mental disease; (3) refused to receive CBT.

Methods

All patients who received single-unit CBT possessed suitable criteria for transplantation, as previously reported^[11]. CB unit selection was based on the total nucleated cell and CD34⁺ cell number among 4/6 to 6/6 HLA-matched units. All cord blood units were obtained through the Chinese Cord Blood Band Network. The conditioning regimen for patients with both AML and ALL was myeloablative conditioning. For patients aged ≤ 14 years or those receiving radiotherapy before transplantation, the conditioning regimen was based on BUCY2 (busulfan 0.8 mg/kg every 6 hours for 4 days, and cyclophosphamide 60 mg/kg daily for 2 days), plus fludarabine (30 mg/m² daily for 4 days). For patients ≥ 14 years old or those with no remission after relapse or primary induction failure, the conditioning regimen was based on total body irradiation cyclophosphamide (total body irradiation 12 Gy given in 4 fractions, and cyclophosphamide 60 mg/kg daily for 2 days), plus high-dose cytarabine (2.0 g/m², every 12 hours for 2 days). Granulocyte colony-stimulating factor was injected subcutaneously at 5 μg/kg daily beginning 6 days after stem cell infusion. The GVHD prophylaxis regimen was a combination of cyclosporine A and mycophenolate mofetil, as previously described^[12]. All laboratory assessments were done as part of clinical care. To evaluate immune reconstitution, peripheral blood samples were obtained and tested 30, 60, 90, 120, 180, and 360 days after UCBT.

Outcomes

The endpoints included GVHD-free/relapse-free survival (GRFS), time to neutrophil and platelet recovery, incidence of non-relapse mortality, relapse, and acute and chronic GVHD (aGVHD and cGVHD, respectively).

Overall survival was calculated from the date of transplantation until death or the last observation alive. The GRFS endpoint was defined as grade III-IV aGVHD, extensive GVHD, relapse, or death. Neutrophil engraftment was defined as absolute neutrophil count > 0.5 × 10⁹/L for 3 consecutive days. Platelet engraftment was defined as platelet count ≥ 20 × 10⁹ L^-1 without platelet transfusion for 7 days^[13]. Both aGVHD and cGVHD were graded according to previously published criteria^[14]. Non-relapse mortality was defined as death without prior relapse. The primary endpoints include GRFS, overall survival, non-relapse mortality and relapse. The secondary endpoints include GVHD and neutrophil and platelet recovery.

Patients had high risk factors at first diagnosis: ALL with adverse cytogenetics or molecular abnormalities [Ph+ chromosome (BCR-ABL positive), hypodiploidy, 11q23 abnormalities (MLL rearrangements), or with a high level of minimal residual disease (1% or more after completion of 6 weeks of induction therapy)]; AML with adverse cytogenetics or molecular abnormalities [t(6;9)(p23;q34.1) or t(v;11q23.3),t(9;22)(q34.1;q11.2), inv(3)(q21.3q26.6),-5 or del(5q),-7,-17/abn(17p), complex karyotype, normal cytogenetics with FLT3-ITD mutation], or prior history of myelodysplastic syndrome^[15-16].

Statistical analysis

Patient- and transplant-related variables were compared between the two patient groups (AML or ALL) using the chi-square statistic for categorical variables and the Mann-Whitney test for continuous variables. Continuous variables were represented by median and categorical variables were represented by proportion. Variables considered in the analysis included patients’ age at transplantation, donor/recipient sex, disease status at transplantation (patients were divided according to the 1^st remission period, the 2^nd remission period, > 2^nd remission, and advanced disease), HLA match, ABO compatibility, type of conditioning (total body irradiation based or BUCY2 based), total nucleated cell dose, and CD34⁺ cell dose. Cumulative incidence functions were used in a competing risk setting, with relapse treated as a competing event to calculate non-relapse mortality probabilities, and with death from any cause as a competing risk for GVHD, engraftment, and relapse. Probabilities of GRFS and overall survival were calculated using Kaplan-Meier estimates. Univariate analyses were performed using Gray’s test for cumulative incidence functions, and log-rank test for GRFS and overall survival^[17]. Baseline variables that were considered clinically relevant or that showed a univariate relationship with outcome were entered into multivariable model. Variables for inclusion were carefully chosen, given the number of events available, to ensure parsimony of the final model. Candidate variables with a P-value < 0.2 on univariate analysis were included in the multivariable model. CD4⁺ T cells, CD8⁺ T cells, regulatory T cells (Tregs), and natural killer cells were compared between AML and ALL using the Mann-Whitney U-test and Kruskal-Wallis test. Statistical analyses were performed with EZR, version 1.37 software (Saitama Medical Center, Jichi Medical University, Japan)^[18], which is a graphical user interface for R version 3.0.2 software (The R Project for Statistical Computing, Vienna, Austria). A value of P < 0.05 was considered statistically significant.

Although the guidelines of the National Comprehensive Cancer Network have highlighted that chronological age is a poor surrogate for suitable therapy, most clinical research has still been based on age groups when considering the biological difference between AML and ALL, which have at different ages of onset. In light of the differences in chemotherapy between AML and ALL, previous studies have not compared the two diseases in one therapy but rather considered them as independent factors with respect to the outcomes of therapy^{[15, 19]}. Even in HSCT therapy, where both AML and ALL with the same risk level have the same conditioning regimens and GVHD prophylaxis, few studies have compared these two diseases in one type of HSCT. Among the various HSCT therapies, CBT is widely used in children with acute leukemia, especially pediatric ALL, which accounts for the largest proportion of pediatric leukemia; however, CBT is not recommended in adults as the first choice because of the limited CB dose^[20]. In contrast, the median age of onset in AML is 67 years, with 54% of patients aged 65 years or older. Thus, there are more patients with AML undergoing transplantation than those with ALL, with CBT representing a considerable proportion of these transplants^[21]. In the present study, grouping by age was used to eliminate the biological difference between AML and ALL. As compared with previous reports on single-unit CBT in AML and ALL, we conducted a systematic study and analysis to answer our research question regarding differences in outcomes of AML and ALL after single-unit CBT using myeloablative conditioning without antithymocyte globulin. For the comparison by grouping of age, we could recommend CBT for the suitable patients with AML or ALL.

Several studies have indicated that pediatric AML and ALL show no difference in survival after CBT^[22-23]. In studies with a restricted age (< 1 year) for the study population, it appears that patients with AML have better survival than those with ALL, with GRFS at 4 years of 40% and 66%, respectively^[9]. The present study showed that childhood acute leukemia had different outcomes after the same type of HSCT, in contrast to previous studies. In this study, the GRFS and overall survival after CBT in children group showed no differences between AML and ALL. Compared with published research with restricted age groups, there were fewer patients aged < 1 year in the age group ≤ 14 years in this study. Thus, the difference in GRFS between pediatric AML and ALL may only for single-unit CBT at a very young age but not for other types of HSCT at other ages.

There are several studies showing delayed engraftment of single-unit CBT, in contrast with other types of HSCT, such as unrelated peripheral blood stem cell transplantation and haploidentical HSCT^[24-26]. However, there are no reported differences in engraftment between AML and ALL using one style of HSCT, especially single-unit CBT. In the present study, AML and ALL also showed no difference in engraftment for CBT, without grouping patients by age. However, in the adolescents and young adults group, the recovery of neutrophils and platelets in patients with AML was better than in those with ALL. Among adolescents and young adults, patients with ALL underwent conditioning with total body irradiation more than in patients with AML. This result was in accordance with previous findings that total body irradiation causes great injury to bone marrow stromal cells and has a high risk of infection^[27]. Their results were also according to the influence of total body irradiation for injury to bone marrow stromal cells. A study indicated that regulatory T-cells play a critical role in preventing autoimmune-mediated thrombocytopenia; in addition, delayed immune reconstitution, including regulatory T-cells, may contribute to the delay of platelet recovery^[28-29]. Our data of immune reconstitution also showed that Treg cells were lower in ALL than in AML. Studies in adult patients with AML and ALL have showed similar results for aGVHD (grade III-IV or grade II-IV) and cGVHD after CBT^[30-32]. For pediatric AML and ALL, a few studies have also reported similar outcomes of aGVHD and cGVHD after CBT^[33-34]. Our data confirmed similar outcomes of aGVHD and cGVHD between AML and ALL in both adult and pediatric patients.

Similar to findings regarding aGVHD and cGVHD between AML and ALL in previous studies, relapse in these diseases is also similar after CBT^{[1, 3, 21]}. In age-restricted studies among adults and children, relapse in patients aged > 50 and < 1 years was similar between AML and ALL^[9]. However, according to our data, relapse for ALL at 2 years was higher than AML in patients without grouping by age. In line with a previous report, our data also showed that there were no differences in relapse between AML and ALL in children. To further explore the higher relapse rate in ALL, we analyzed immune reconstitution after CBT. Our data suggested that the CD8⁺ T cell dose in AML for the age group 15-39 years was higher than that in ALL at 4 months after CBT. One study mentioned that early CD8⁺ recovery was associated with low risk of relapse and more severe GVHD^[35]. In this study, the CD8⁺ cell count was higher in patients with ALL, who had lower relapse rates. CD8⁺ T cells are also considered to be related to aGVHD; in the present study, we believe that the double negative T cell levels might be related to cGVHD^[36-37]. In continuous monitoring, we detected that there was no difference between AML and ALL with respect to immune reconstitution, after 1 year. Therefore, patients with AML or ALL both had decreased relapse with time.

For transplant recipients, adolescents and young adults are a special group of patients, as the disease occurs early in life, in contrast to older adults, resulting in a high burden of HSCT sequelae^[38]. Earlier studies indicated that between adolescents and young adults and pediatric patients, inferior outcomes were observed with older age^[39-40]. Our research showed similar results in non-relapse mortality between adolescents and young adults and pediatric patients with AML. According to our analysis, only HLA compatibility showed this difference of non-relapse mortality. Adolescents and young adults are characterized by plenty of maturing, hormonal, and psychosocial challenges, as well as developmental processes^[41]. Thus, we believe that nearly all components of meaningful life experiences, which were not included in our research, are affected and are sensitive with respect to outcomes of HSCT.

This study has some limitations. First, this was a retrospective study with inherent biases and a lack of randomization. Second, the sample size was relatively small with grouping by age and weight, especially in older patients. To avoid this limitation, we did not analyze some outcomes in older groups. Therefore, additional studies are necessary to confirm the outcomes of CBT for older patients with AML and ALL.

In summary, we report the outcomes of AML and ALL after CBT using myeloablative conditioning without antithymocyte globulin, according to age groups. The patients with AML showed better outcomes of neutrophil and platelet recovery and relapse than adolescents and young adults in ALL. However, there were no differences between AML and ALL in outcomes after CBT with or without grouping by age, especially with respect to overall survival and GRFS. On the basis of this study, adults with ALL have outcomes of CBT that are as good as those of patients with AML. Therefore, CBT could be a decent choice for adults who do not have a suitable donor. 

中国组织工程研究杂志出版内容重点：干细胞；骨髓干细胞；造血干细胞；脂肪干细胞；肿瘤干细胞；胚胎干细胞；脐带脐血干细胞；干细胞诱导；干细胞分化；组织工程

BACKGROUND: Umbilical cord blood hematopoietic stem cell transplantation is more and more widely used as a radical treatment for acute leukemia, but its therapeutic effect in different leukemias has not been compared. By comparing the efficacy of diseases, it can guide different patients to choose the transplantation method.

中国组织工程研究杂志出版内容重点：干细胞；骨髓干细胞；造血干细胞；脂肪干细胞；肿瘤干细胞；胚胎干细胞；脐带脐血干细胞；干细胞诱导；干细胞分化；组织工程#br#