Patients Characteristics
A total of 33 AML patients included 13 males and 20 females, with a median age of 42 years (10–57 years) and 15 MDS-EB patients included 11 males and 4 females, with a median age of 51 years (4–60 years). The median follow-up time was 26 (0.7–56) months after HSCT. Patient characteristics are shown in Table 1. Of these 33 AML patients, 31 patients had gotten CR after chemotherapy, and 2 patients had gotten NR after 3 courses of chemotherapy. And 5 MDS-EB patients receiving chemotherapy including decitabine had gotten CR pre-transplantation. All patients had neutrophil engraftment, and 39 patients had platelet engraftment. Of the 33 patients with AML, 7 patients (21.2%) died of TRM and 6 patients (18.2%) underwent hematological relapse who died ultimately. The median hematological relapse time was 4.8 months (range 4–9 months) after HSCT in 6 relapsed patients. Of the 15 patients with MDS-EB, 2 patients (13.3%) died of infection. In addition, all enrolled patients had a 3-year CIR of 13.7%±5.2%, 3-year OS of 67.8%±6.9%, 3-year DFS of 68.1%±6.8% and 3-year TRM of 20.3%±6.1% (Fig. 1).
Table 1
Characteristics of acute myeloid leukemia and myelodysplastic syndrome patients.
Characteristic
|
AML
N = 33
|
MDS-EB1/2
N = 15
|
Median age at allo-HCT, years (range)
|
42 (10–57)
|
51 (4–60)
|
Gender, n (%)
|
|
|
Male
|
13 (39.4%)
|
11 (73.3%)
|
Female
|
20 (60.6%)
|
4 (26.7%)
|
Chromosome normal, n (%)
|
23 (69.7%)
|
9 (60.0%)
|
FLT3-ITD mutation, n (%)
|
|
|
Yes
|
10 (30.3%)
|
0
|
No
|
23 (69.7%)
|
15 (100%)
|
NPM1 positive, n (%)
|
0
|
0
|
Median WT1 expression level at initial diagnosis
|
25.25 (0.23–83.20)
|
18.80 (1.40–53.50)
|
No remission before transplant, n (%)
|
2 (6.1%)
|
1 (6.7%)
|
Donor type, n (%)
|
|
|
HLA-matched sibling
|
7(21.2%)
|
5(33.3%)
|
Haploidentical
|
26(78.8%)
|
10(66.7%)
|
ABO blood type match, n (%)
|
|
|
Compatible
|
17 (51.5%)
|
7 (46.7%)
|
Incompatible
|
16 (48.5%)
|
8 (53.3%)
|
Conditioning regimen, n (%)
|
|
|
Chemotherapy based
|
33 (100%)
|
15 (100%)
|
TBI based
|
0
|
0
|
Cell compositions in allografts
|
|
|
Median MNC, ×108/kg (range)
|
7.82 (6.04–10.86)
|
8.54 (6.10-10.86)
|
Median CD34 + count, ×106/kg (range)
|
2.32 (0.27–6.67)
|
1.89 (0.84–5.34)
|
Granulocyte engraftment time, day(range)
|
13 (8–25)
|
13 (11–19)
|
Platelet engraftment time, day (range)
|
14 (10–74)
|
13 (10–53)
|
II–IV°aGVHD
|
8 (24.2%)
|
1 (6.7%)
|
aGVHD
|
18 (54.5%)
|
4 (26.7%)
|
DLI after transplant, n (%)
|
|
|
For relapse prevention
|
2 (6.1%)
|
0
|
For intervention
|
4 (12.1%)
|
2 (13.3%)
|
Prognosis, n (%)
|
|
|
Relapse
|
6 (18.2%)
|
0
|
Treatment-related death
|
7 (21.2%)
|
2 (13.3%)
|
Relapse death
|
6 (18.2%)
|
0
|
AML, acute myeloid leukemia; MDS, myelodysplastic syndrome; HLA, human leukocyte antigen; TBI, total body irradiation; MNC, mononuclear cell ;aGVHD, acute graft versus host disease; DLI, donor lymphocyte infusion. |
Dynamic changes of MLL-PTD before and after transplantation
Observing the changes in the expression level of MLL-PTD at different time points peri-transplantation is helpful to analyze the stability of MLL-PTD as an MRD indicator in the transplantation system. Our results showed that the MLL-PTD level before transplantation was significantly lower than that at the initial diagnosis, but there were still 37 cases were MLL-PTD positive before transplantation, and 33 of 37 cases became negative within post-transplant 1 month. However, during our follow-up period, 25 cases eventually occurred post-transplant MLL-PTD positive. The median MLL-PTD level in all enrolled patients was decreased by around 35 folds after transplantation compared with that of pre-transplant CR status and was similar to the healthy controls (Table 2). Furthermore, among the 6 relapsed patients after transplantation, 3 of them maintained MLL-PTD at the healthy level (< 0.08%) within a month after transplantation. But before relapse, the MLL-PTD level of these 3 patients gradually increased (> 0.08%) and reached the highest level at the time of relapse. The MLL-PTD level of the other 3 relapsed patients continuously remained > 0.08% after transplantation, and the MLL-PTD levels of these 3 patients suddenly increased by hundreds of times before relapse.
Table 2
Comparison of MLL-PTD and WT1 at the initial diagnosed and peri-transplant patients.
|
MLL-PTD > 0.08% (n/total tests, positive rate)
|
Median level of MLL-PTD > 0.08% (range, %)
|
Median level of MLL-PTD (range, %)
|
WT1 > 0.6%
(n/total tests)
|
Median level of WT1 > 0.6% (range, %)
|
P value
(MLL-PTD + vs. WT1+)
|
The initial diagnosis
|
48/48(100%)
|
30.30 (1.20–631.00)
|
30.30 (1.20–631.00)
|
44/47(93.6%)
|
26.20 (0.82–83.20)
|
0.233
|
Pre-transplantation
|
37/48(68.8%)
|
6.10 (0.10-414.10)
|
1.70 (0.017–414.10)
|
28/47(59.6%)
|
6.20 (0.88–53.50)
|
0.351
|
Post-transplantation
|
|
|
|
|
|
|
+ 1 month
|
8/43(18.6%)
|
0.115 (0.083–0.73)
|
0.046 (0.01–0.73)
|
1/46(2.2%)
|
0.82
|
0.027
|
+ 2 month
|
12/44(27.3%)
|
0.21 (0.09–0.82)
|
0.047 (0-0.82)
|
3/45(6.7%)
|
0.86 (0.74–2.4)
|
0.009
|
+ 3 month *
|
13/45(28.9%)
|
0.28 (0.086–104.70)
|
0.05 (0-104.70)
|
6/46(13.0%)
|
1.50 (0.75–32.70)
|
0.063
|
+ 4.5 month *
|
8/38(21.1%)
|
1.30 (0.082–55.30)
|
0.0515 (0-55.30)
|
9/39(23.1%)
|
3.90 (0.81–44.10)
|
0.524
|
+ 6 month*
|
11/39(28.2%)
|
1.40 (0.096–101.30)
|
0.053 (0.015–101.30)
|
12/39(30.8%)
|
1.30 (0.60–80.90)
|
0.500
|
+ 9 month*
|
5/27(18.5%)
|
0.09 (0.08–0.11)
|
0.0445 (0–1.00)
|
3/34(8.8%)
|
0.71 (0.63–0.74)
|
0.231
|
+ 12 month
|
1/30(3.3%)
|
0.45
|
0.049 (0-0.45)
|
5/32(15.6%)
|
0.88 (0.72-1.00)
|
0.113
|
*Patients underwent hematological relapse at that time point. |
The effect of MLL-PTD level before and after transplantation on prognosis
Having analyzed the dynamic changes above which peri-transplant MLL-PTD can stably reflect the disease state we next studied the optimal threshold of post-transplant MLL-PTD for relapse. Our previous study shows that patients with MLL-PTD/ABL ≥ 1% based on initial diagnosis have a poor clinical prognosis [13]. In order to explore whether MLL-PTD could be used as a MRD marker after transplantation, we performed a receiver operating characteristic (ROC) with the highest expression level of post-transplant MLL-PTD before hematological relapse in all patients to determine the optimal cut-off value to predict relapse. The area under the ROC curve value was 0.977 (P < 0.001, Fig. 2A). The optimal cut-off value was MLL-PTD/ABL = 1.0%. And as shown in Fig. 2B, most post-transplant patients with MLL-PTD maintained a low level of expression, only 8 patients had MLL-PTD ≥ 1%, and 6 of the 8 patients eventually relapsed, which also implied the importance of MLL-PTD ≥ 1% in predicting relapse after transplantation. Based on the optimal cut-off value, we divided the post-transplant patients into two groups of MLL-PTD/ABL < 1% and MLL-PTD/ABL ≥ 1% to analyzed the prognostic difference. Our study found that the group of MLL-PTD/ABL ≥ 1.0% had higher 3-year CIR (75%±15.3% vs. 0%, P < 0.001, Fig. 3A), and lower 3-year OS (25.0 ± 15.3% vs. 80.7%±6.6%, P < 0.001, Fig. 3B) and 3-year DFS (25.0 ± 15.3% vs. 80.7%±6.6%, P < 0.001, Fig. 3C) compared with that of group of MLL-PTD/ABL < 1%. However, there was no statistical difference between the two groups in TRM (P > 0.05, Fig. 3D).
Both at the initial diagnosis and post-transplantation, it was analyzed that MLL-PTD = 1% was the optimal cut-off value, which implied that MLL-PTD/ABL = 1% was of important value in predicting prognosis. Therefore, we further analyzed whether MLL-PTD/ABL ≥ 1% before transplantation also indicated a poor prognosis after transplantation. However, our results showed that there was no statistical difference in prognosis between the MLL-PTD/ABL ≥ 1% and MLL-PTD/ABL < 1% group based on the level of MLL-PTD before transplantation (All P > 0.05, Fig. 4A-4D), but the group of MLL-PTD/ABL ≥ 1% tended to have lower OS (P = 0.202, Fig. 4B), DFS (P = 0.202, Fig. 4C), and have a higher TRM(P = 0.105, Fig. 4D) compared with that of MLL-PTD/ABL < 1% group .
Factors affecting the prognosis of transplant patients with MLL-PTD
Factors affecting the prognosis were analyzed, including transplantation age, gender, disease type, donor type, blood type compatibility (Table 3). There was no statistical difference in TRM (P = 0.675), CIR (P = 0.115), DFS (P = 0.151) and OS (P = 0.157) between AML and MDS-EB. Among the 12 patients who received MSDT, 2 (16.7%) patients underwent hematological relapse both at 5 months after HSCT, and 1 patient died of pneumonia at 5.5 months. Among 36 patients who received haplo-HSCT, 4 patients (11.1%) underwent hematological relapse at a median of 4.5 months (range, 4–9 months) after HSCT, and 8 patients (22.2%) died due to TRM at a median of 5.3 months (range, 0.7–17.5 months). Based on the results of the analysis, it seemed that patients who received haplo-HSCT could achieve comparable outcomes compared to those who underwent MSDT (TRM: P = 0.271; CIR: P = 0.653; DFS: P = 0.544; OS: P = 0.560). The factor analysis of MLL-PTD level before and after transplantation showed that there was no statistical difference in pre-transplant MLL-PTD level. And post-transplant group of MLL-PTD/ABL ≥ 1% had a higher CIR, a lower OS and a lower DFS than that of group of MLL-PTD/ABL < 1% (all P < 0.001). In addition, other factors such as age, pre-transplant FCM, WT1 status and prophylactic DLI have no significant impact on prognosis. The ABO blood type and FLT3-ITD mutation at first diagnosis were important risk factors of CIR and OS after transplantation, respectively. Incompatible ABO blood type indicated a higher CIR than that of compatible ABO blood type, and patients with FLT3-ITD mutation had a low OS than that of without FLT3-ITD (Table 3).
Table 3. Univariate analysis of the variables affecting hematological TRM, CIR, DFS and OS in patients with MLL-PTD after allo-HSCT
Variables
|
Number
(n, %)
|
P value
|
TRM
|
CIR
|
DFS
|
OS
|
Age of recipient
|
|
0.965
|
0.291
|
0.410
|
0.442
|
<50 years
|
31(64.6%)
|
|
|
|
|
≥50 years
|
17(35.4%)
|
|
|
|
|
Underlying disease
|
|
0.675
|
0.115
|
0.151
|
0.157
|
AML
|
33(68.8%)
|
|
|
|
|
MDS-EB1/2
|
15(31.2%)
|
|
|
|
|
ABO compatibility
|
|
0.264
|
0.009
|
0.38
|
0.484
|
Compatible
|
24(50.0%)
|
|
|
|
|
Incompatible
|
24(50.0%)
|
|
|
|
|
Donor type
|
|
0.271
|
0.653
|
0.544
|
0.560
|
HLA-matched sibling
|
12(25.0%)
|
|
|
|
|
Haploidentical
|
36(75.0%)
|
|
|
|
|
Prophylactic DLI
|
2(4.2%)
|
0.325
|
0.591
|
0.735
|
0.702
|
FLT3-ITD positive
|
10(20.8%)
|
0.067
|
0.868
|
0.068
|
0.041
|
Pre-transplantation FCM
|
|
0.056
|
0.504
|
0.291
|
0.232
|
Negative
|
23(47.9%)
|
|
|
|
|
Positive
|
25(52.1%)
|
|
|
|
|
Pre-transplantation WT1
|
|
0.339
|
0.166
|
0.843
|
0.854
|
WT1<0.6%
|
19(40.4%)
|
|
|
|
|
WT1≥0.6%
|
28(59.6%)
|
|
|
|
|
Pre-transplantation MLL-PTD
|
|
0.105
|
0.967
|
0.202
|
0.202
|
MLL-PTD/ABL≥1.0%
|
25(52.1%)
|
|
|
|
|
MLL-PTD/ABL<1.0%
|
23(47.9%)
|
|
|
|
|
Post-transplantation MLL-PTD
|
|
0.277
|
<0.001
|
<0.001
|
<0.001
|
MLL-PTD/ABL≥1.0%
|
8(16.7%)
|
|
|
|
|
MLL-PTD/ABL<1.0%
|
38(79.2%)
|
|
|
|
|
TRM, treatment-associated mortality; CIR, cumulative incidence of relapse; DFS, disease-free survival; OS, overall survival; HLA, human leukocyte antigen; allo-HSCT, allogeneic hematopoietic stem cell transplantation; DLI, donor lymphocyte infusion; MLL-PTD, mixed lineage leukemia-partial tandem duplication; AML, acute myeloid leukemia; MDS, myelodysplastic syndrome
Comparison of MLL-PTD and other MRD parameters
After transplantation, 8 patients were detected MLL-PTD/ABL ≥ 1.0% at a median of 3 months. Of the 8 patients, 7 patients were simultaneously (5 patients) or subsequently (2 patients) MRD positive detected by FCM at a median of 4.25 months (range,3–12 months), and 6 patients ultimately progressed to hematological relapse at a median of 2 months (range, 0.25–6 months) from the first time MLL-PTD/ABL ≥ 1.0%, half of whom receiving chemotherapy plus DLI. Finally, 2 patients receiving chemotherapy plus DLI became MRD negative gradually.
WT1 has been confirmed in previous studies to be an effective indicator of MRD monitoring and implementing interventions [24]. In order to analyze the specificity and sensitivity of MLL-PTD compared with WT1, we showed in Table 2 the dynamic changes of expression of MLL-PTD and WT1 at the initial diagnosis and different time points before and after transplantation. All 6 relapsed patients were detected MLL-PTD positive prior to relapse, while only 4 patients were detected positive for WT1. As shown in Table 2, the expression levels of MLL-PTD and WT1 both changed with the tumor burden. However, within post-transplant 3 months, MLL-PTD seemed be more sensitive than WT1 for MRD monitoring (P+ 1 month=0.027; P+ 2 month=0.009; P+ 3 month=0.063).