3.1 Patient characteristics
A total of 79 patients who were eligible to receive HSCT were enrolled in the study. All patients were suggested to receive an additional third-party UCB infusion, and 55 patients provided informed consent. After searching the suitable UCB units as a previous scheme in Liaoning and Shandong UCB banks, 53 patients having suitable UCB were recruited into the UCB-haplo group. Twenty-four patients did not sign the consent to receive UCB, and 2 patients had no suitable UCB in the haplo group (Fig. 1). The patients’ characteristics are summarized in Table 1.
3.2 UCB and haploid graft characteristics
Following the previously described protocol, UCB units were chosen by the same physician and rechecked by another transplantation group physician. The characteristics of UCB units and haploid grafts are listed in Table 2. In addition, the quantities of MNC and CD34+ cells in haploid grafts were not significantly different between the UCB-haplo group and the haplo group. For HLA matching between the UCB unit and recipient, 10 units were 9/10 matched, 26 units were 8/10 matched, and 17 units were 7/10 matched. Before UCB infusion, an anti-allergy regimen was performed. There was no obvious transfusion reaction observed.
3.3 Engraftment
All surviving patients underwent chimerism analysis at Day 30 after HSCT, and they all achieved full donor chimerism. Chimerism analyses in these patients were continued regularly until disease relapse. During follow-up, two patients experienced mixed chimerism. One patient was found to have mixed chimerism at the six-month visit with a normal range of routine blood tests and died at Day 235 after HSCT because of a serious fungal infection. The other patient was found to have mixed chimerism at Day 60 after HSCT and turned to full donor chimerism at Day 90 after HSCT.
The day of neutrophil and platelet engraftment was not significantly different between the UCB-haplo group and the haplo group. The median day of neutrophil engraftment was at Day 12 (range, 10-24) and Day 13 (range, 10-42) for the UCB-haplo group and haplo group, respectively (p=0.349). The cumulative incidence of neutrophil engraftment on Day 30 was 100% in the UCB-haplo group and 96% (95% CI, 72.7%-99.4%) in the haplo group (P=0.52). Meanwhile, the median day of platelet engraftment was at Day 14 (range, 9-69) and Day 13 (range, 8-96) for the UCB-haplo group and haplo group, respectively (p=0.974). The cumulative incidence of platelet engraftment on Day 100 was 100% in both the UCB-haplo group and haplo group (P=0.55).
3.4 GVHD
Both aGVHD and cGVHD were considered in the present study. In the UCB-haplo group, the 100-day cumulative incidences of grade II-IV aGVHD and grade III-IV aGVHD were 24.53% (95% CI, 12.01-35.27%) and 5.67% (95% CI, 0-11.68%), respectively. However, in the haplo group, the 100-day cumulative incidences of grade II-IV aGVHD and grade III-IV aGVHD were 15.38% (95% CI, 0.31-28.18%) and 7.69% (95% CI, 0-17.39%), respectively. There was no significant difference between the two groups (p=0.36 and 0.73, respectively). The 1-year cumulative incidence rates of cGVHD in both the UCB-haplo group and the haplo group were 30.19% (95% CI 16.67-41.52%) and 38.46% (95% CI 16.61-54.59%), respectively. There was also no significant difference between the two groups (p=0.45).
3.5 CMV and EBV infection
The 100-day cumulative incidence of CMV viremia was 47.17% (95% CI, 31.87-59.04%) in the UCB-haplo group versus 50.00% (95% CI, 26.56-65.96%) in the haplo group (p=0.78).
The 100-day cumulative incidence of EBV viremia was 39.62% (95% CI, 24.9-51.45%) in the UCB-haplo group versus 19.23% (95% CI, 2.57-33.04%) in the haplo group (p=0.062).
3.6 OS, PFS, TRM and relapse rate
The probability of OS in the UCB-haplo and haplo groups was 79.92% (95% CI, 69.39-92.04%) and 61.54% (95% CI, 45.41-83.39%), respectively (p=0.035). (Figure 2) PFS in the UCB-haplo and haplo groups was 74.92% (95% CI, 63.25-88.74%) and 53.85% (95% CI, 37.72-76.86%), respectively (p=0.011). (Figure 2)
The cumulative incidence of TRM was 18.13% (95% CI, 6.47–28.34%) and 35.38% (95% CI, 13.78-51.58%) in the UCB-haplo and haplo groups, respectively (p=0.045). (Figure 2) The cumulative incidence of relapse was 8.10% (95% CI, 0.00–16.87%) and 28.57% (95% CI, 6.38-45.50%) in the UCB-haplo and haplo groups, respectively (p=0.004). (Figure 2)
The probability of PFS excluding AA in the UCB-haplo and haplo groups was 66.72% (95% CI, 51.76-85.99%) and 50.00% (95% CI, 32.92-75.94%), respectively (p=0.049). The cumulative incidence of relapse excluding AA patients was 12.36% (95% CI, 0.00–25.39%) and 29.41% (95% CI, 4.06-48.06%) in the UCB-haplo and haplo groups, respectively (p=0.029).
3.7 Univariate analysis
Univariate analysis was performed to determine the factors for predicting survival and relapse. The basic characteristics of patients and donors were used to identify predictive factors, and unrelated UCB, GVHD and infection were also included. The results showed that older age, poor disease status, high EBMT risk score, accompanying cGVHD, infection and the absence of a UCB unit were associated with poor OS and PFS as measured by univariate analysis (p<0.10, shown in Table 3). Per univariate analysis, disease status, EBMT risk score, cGVHD, infection and the combination of UCB were also related to TRM (p<0.10, shown in Table 3). However, poor disease status, high EBMT risk score and no combination of UCB were related to relapse (p<0.10, shown in Table 3).
3.8 Multivariate analysis
In multivariate analysis (Table 4), the combination of UCB units (HR 0.331; 95% CI, 0.130-0.843; P = 0.020) was an independent factor for improving OS. Infection (HR 2.964, 95% CI, 1.172-7.496; P = 0.022) and cGVHD (HR 0.199; 95% CI, 0.054-0.729; P = 0.015) were independent factors associated with poor OS. Meanwhile, the combination of UCB units (HR 0.338; 95% CI, 0.144-0.795; P = 0.013) was an independent factor for improving PFS. However, cGVHD (HR 0.232; 95% CI, 0.074-0.72; P = 0.012) was an independent factor associated with poor PFS. For TRM, the combination of UCB units (HR 0.329; 95% CI, 0.121-0.889; P = 0.028) was an independent factor for reducing TRM. Infection (HR 4.104, 95% CI, 1.506-11.186; P = 0.006) and cGVHD (HR 0.228; 95% CI, 0.061-0.850; P = 0.028) were independent factors associated with increased TRM. The only independent factor for relapse was the combination of UCB units (HR 0.243; 95% CI, 0.061-0.973; P = 0.046), which could reduce relapse.