Introduction
Allogeneic hematopoietic stem cell transplantation (allo-HSCT) might be the only curative approach for many children with myelodysplastic syndrome (MDS). It is routinely implemented to patients with advanced MDS (aMDS) (including refractory anemia with excessive blasts [RAEB] and RAEB in transformation [RAEB-t]) or patients with refractory cytopenia of childhood (RCC) accompanied with monosomy 7, complex karyotype, severe neutropenia or transfusion dependenc e[
1,
2]. Watch-and-wait strategy or immunosuppression therapy (IST) may be a choice for RCC patients with hypocellular bone marrow (BM) and absence of monosomy 7 and complex karyotype. However, a relevant proportion of those patients still need allo-HSCT subsequently for non-response or relaps e[
3]. The recent update of the European Working Group of MDS in childhood (EWOG-MDS) data showed that the survival outcomes of patients transplanted from either a human leukocyte antigen identical (HLA-identical) sibling or an unrelated donor (UD) matched for 9/10 or 10/10 HLA-loci might be almost comparabl e[
4]. However, cord blood transplantation (CBT) resulted in survival rates below 30-60 %[
5‐
7]. As for haploidentical transplantation, data is far more limited, remaining identified. Interestingly, the 5-year overall survival rate (OS) of HSCT from haploidentical family donors for pediatric patients with MDS was as high as 86% in a recent Korean cohor t[
8]. Generally, there is still a paucity of data to inform the best transplant type for pediatric MDS.
As for the pre-transplant period for aMDS, a diversity of therapy strategies like intensive chemotherapy, AML-type induction chemotherapy, minimally myelosuppressive regimen (MMR) and DNA methyltransferase (DNMT) inhibitors has been investigate d[
9,
10]. Intensive chemotherapy is not generally recommended due to showing no survival benefi t[
11]. The acute myeloid leukemia type (AML-type) induction chemotherapy is controversial because of its somewhat severe toxicity and considerable mortalit y[
10,
11]. The debut of low-dose induced remission treatment (low dose of cytarabine and anthracycline in conjunction with granulocyte colony stimulating factor [G-CSF]) was reported in 1995, being applied among aged patients with myeloid tumor and yielding certain efficac y[
12]. Then, it has been continuously improved and demonstrated to be efficacious during the past decades mainly among adult MDS/AML population s[
13,
14]. Intriguingly, pediatric AML patients receiving MMR (one-tenth of standard dose of cytarabine, one-half dose of anthracycline in conjunction with G-CSF) showed similar outcomes and mutation clearance levels, but significantly lower toxicity compared with those receiving standard chemotherapy in our cente r[
15,
16].
Hypermethylation of critical genes was revealed in adult and childhood MDS, considered one of the disease’s driving alteration s[
9,
17,
18]. In addition, hypermethylation of the promoters of various genes was associated with unfavorable prognosis in MDS, and the strategy of adopting DNA-hypomethylating agents including azacitidine (AZA) and decitabine (DAC) combination therapy is appealing for MD S[
19,
20]. It has been widely recognized that low-dose DNA-hypomethylating agents could improve the quality of life and prolong survival to a significant extent for old people, especially for those unfit for allo-HSCT or intensive chemotherap y[
21]. However, the role of DNA-hypomethylating agents in the treatment of childhood MDS is scarc e[
22,
23].
Considering the possible advantages of disease control with good tolerability during HSCT preparation, improved antitumoral alloimmunity, reduced risk of relapse, and so on,[
24‐
26] we have upheld a scientific hypothesis that decitabine-combined MMR strategy (DAC + MMR) bridged allo-HSCT may be a feasible way for pediatric aMDS patients with low toxicity and high efficiency. Here, we present retrospective data on the 28 children with de novo MDS who underwent allo-HSCT during the past decade in our single center. The clinical features, chemotherapy regimens, transplant characteristics, outcomes, adverse events and complications were investigated and analyzed.
Patients and methods
Patient population
A total of 28 pediatric MDS patients hospitalized and receiving allo-HSCT from January 2011 to December 2020 at our single center were finally enrolled in this study. Both experienced hematologists and pathologists reviewed the diagnosis of all patients. They were newly diagnosed as de novo MDS according to pediatric modification of the World Health Organization (WHO) classification. According to the current recommendations, they were categorized as RCC, RAEB and RAEB-t [
27,
28]. Following the proposed categorization by
Hasle et al., patients with RCC were termed as low-grade MDS while those with RAEB or with RAEB-t were termed as advanced MDS (aMDS) [
29].
The inclusion criteria were as follows: (1) younger than 14 years of age at disease onset; (2) newly diagnosed as de novo MDS; (3) not Down syndrome (DS)-related MDS; (4) receiving allo-HSCT after diagnosis. Patients who developed AML at any time before transplantation were excluded. Cytogenetic analysis of BM cells was performed for all of the patients.
The indications for allo-HSCT among MDS patients were: (1) RCC patients with monosomy 7, 7q deletion or complex karyotype; (2) RCC patients with severe neutropenia or transfusion dependence; (3) aMDS patients.
Ethical statement
This retrospective study was authorized by the local ethical committee of Children’s Hospital of Soochow University. The written informed consents were obtained from the patients’ parents or legal guardians. The study is carried out in accordance with the Declaration of Helsinki.
Chemotherapy
The AML-type induction chemotherapy was similar to the protocol of AML induction remission therapy used in our cente r[
15,
16]. The decitabine-combined minimally myelosuppressive regimen (DAC + MMR) included three subtypes of regimens. One subtype was “DAC + MAG”, which contained decitabine (20 mg/m
2 once a day intravenously from the first day to the 5th day), cytarabine (10 mg/m
2 every 12 hours subcutaneously from the 6th to 15th day), mitoxantrone (5 mg/m
2 once a day intravenously for the 6th, 8th and 10th day) and G-CSF (5 μg/m
2 once a day subcutaneously from the 6th to 15th day). One was aliased as “DAC + HAG”, which contained decitabine, cytarabine and G-CSF with the same usage as above and homoharringtonine (1 mg/m
2 once a day intravenously from the 6th to 12th day). And the third one contained decitabine, cytarabine, and G-CSF with the same usage as above, and idarubicin (5 mg/m
2 once a day intravenously from the 6th to 8th day) was abbreviated as “DAC + IDAG”. Additionally, sole decitabine treatment prior to transplantation performed as decitabine at 20 mg/m
2 once a day intravenously for five consecutive days was applied for some aMDS patients with BM blasts slightly higher than 5%.
Transplantation
The conditioning regimens included myeloablative conditioning (MAC) and reduced-intensity conditioning (RIC). All the regimens were busulfan and cyclophosphamide based (Bu + Cy) or fludarabine and busulfan based (Flu+Bu). The types of transplantation included HLA-identical transplantation (containing sibling donor allo-HSCT [sib-HSCT] and unrelated matched HSCT), haploidentical transplantation, and cord blood transplantation (CBT). The graft-versus-host disease (GVHD) prophylaxis contained calcineurin inhibitors (cyclosporine A or tacrolimus), mycophenolate mofetil, as well as short-term methotrexate.
Evaluation and criterion
The neutropenia was defined as absolute neutrophil count (ANC) < 1.5*10
9 /L and severe neutropenia was ANC < 0.5*10
9 /L. The thrombocytopenia was defined as platelet count (Plt) < 100*10
9 /L, and severe thrombocytopenia was Plt < 20*10
9 /L and/or clinical need for platelet transfusion. The response to treatment was assessed by reference to the International Working Group (IWG) response criteria in myelodysplasia [
30]. Marrow complete remission (mCR) referred to the achievement of marrow blasts ≤5% with or without improved cytopenias. Adverse events of administered treatments were graded by using the common terminology criteria of adverse events score (CTCAE) (version 4.0). Graft failure (the primary) was defined as ANC that did not maintain sustained engraftment (> 0.5*10
9 /L) within 28 days post-transplantation. The granulocyte engraftment was defined as ANC ≥0.5*10
9/L for three consecutive days. The platelet engraftment was defined as Plt ≥20*10
9/L for seven consecutive days without platelet transfusion support. The acute and chronic graft-versus-host disease (GVHD) were graded based on traditional criteria [
31,
32].
Follow-up
All the patients were followed up every month and the follow-up endpoint was August 31, 2021. The overall survival time (OS) was calculated from the date of first diagnosis to the date of death or last follow-up. The events included death, relapse, graft failure, severe complications (acute renal failure, for instance) and secondary tumor (progression to AML, for instance) and the event-free survival time (EFS) was defined as survival without those events. Relapse was defined as morphological evidence of disease in BM or recurrence and sustained pre-transplant chromosomal abnormalities. The relapse-free survival (RFS) time was calculated.
Statistical analysis
The continuous variables with normal distribution were expressed as mean and standard deviation, while variables with skewed distribution were expressed as median and range. The categorical variables were described as number and percentage. The independent-samples T test was used to assess normal distributional variables. The Mann-Whitney U test and Kruskal-Wallis H test were used to assess skewed distributional variables, as appropriate. The categorical variables were analyzed using Chi square or Fisher’s Exact Test, as appropriate. The Kaplan-Meier methods were used to describe survival functions and the log-rank test was used to compare the survival curves. A Cox’s proportional hazards regression model was used to determine the significance of risk factors for the outcomes. Factors with at least P-value< 0.10 in the univariate analysis were included in the model. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated. SPSS 26.0 software was employed for data processing. GraphPad Prism 8.0.2 software was served as the tool for results visualization. Two-tailed P-values< 0.05 were considered statistically significant.
Discussion
Here, we reported a cohort of children with MDS who underwent allo-HSCT over the past decade at our single center. To the best of our knowledge, this may be the latest research in China for systematically reviewing a certain size of cohort regarding pediatric MDS with transplantation and is also the first domestic study in China that reported the experience of decitabine-combined minimally myelosuppressive regimen prior to allo-HSCT for pediatric aMDS.
Pediatric MDS is a heterogeneous group of clonal disorder accounting for less than 5% of childhood hematological malignancies. The morphology, cytogenetics and therapy approaches would profoundly influence the survival outcome s[
33]. It is recognized that patients with abnormal karyotype such as monosomy 7 or complex karyotype are more likely to progress to advanced disease and have poor outcome s[
34,
35]. In the present study, ten patients with abnormal karyotypes had a significant low survival rate compared to 18 patients whose karyotypes were normal. However, the cytogenetic data is far more limited with great heterogeneity which should be carefully interpreted. Recently, with the increased access to gene mutation landscape, genetic counseling for both patients and their families would affect pediatric MDS’s clinical diagnosis and therapeutic decision s[
36‐
38]. The gene mutation assay was performed among the 16 patients of this cohort and 14 of them were verified to carry different gene mutations (Supplementary Table
1). It will be a great challenge for pediatric hematologists further to explore the underlying conditions and their hematopoietic impacts.
As for treatment strategy, it is widely accepted that allogeneic HSCT is the only curative treatment for pediatric MD S[
5,
39]. Especially, high-risk subtype of MDS is recommended to receive an early transplantation. Allo-HSCT for pediatric MDS has been adeptly mastered during the past decade in our center. The 4-year OS as high as 71.4 ± 8.5% for the total cohort, 85.7 ± 13.2% and 66.7 ± 10.3% for low-grade and advanced MDS respectively are revealed in our study, consistent with that of recent reports ranging from 30 to 80 %[
1,
5,
40,
41]. It is reported that allo-HSCT from a matched related or unrelated donor offers a superior survival probability for pediatric MD S[
4]. The data of the seven patients with HLA-identical transplantation in our center confirmed this conclusion again. Our preliminary data showed that the 4-year OS of haploidentical HSCT was 72.2 ± 10.6%, which indicated that haploidentical HSCT would be a feasible alternative among childhood MDS for an urgent need of transplantation. Consistently, a Korean group reviewed 36 pediatric patients with MDS who proceeded to haploidentical HSCT (
n = 9) or HLA-identical HSCT (
n = 24) or CBT (
n = 3 )[
8]. Intriguingly, the OS of HSCT from haploidentical family donors was comparable with that from HLA-identical donors (86% versus 79%,
P = 0.625 )[
8]. With the theoretic advantages, including low incidence of acute and chronic GVHD, despite multiple HLA mismatching and so on, cord blood has been considered as an attractive source for transplantatio n[
42,
43]. However, in our center, all the three patients with CBT in the cohort died soon after transplantation, leading to no obvious benefit regarding overall survival. In the future, more data of CBT will be needed to draw certain conclusions. The cumulative incidence of transplantation-related mortality (TRM) for the total cohort was 28.6 ± 8.5%. Acute GVHD is a serious transplant complication that contributes TRM after allo-HSC T[
43]. In the risk factor analysis for OS, grade III-IV aGVHD was associated with higher risk of mortality and should be prevented.
It is challenging and time-consuming for donor searching and HSCT preparation. Therefore, the disease should be controlled through a bridging treatment based on risk stratification. What has been reached as a common consensus is that conventional chemotherapy is dubious, especially for high-risk MDS. The advent of epigenetic treatment options for myeloid disorders has led to the combination concepts, and their integration with transplantation already shows a reliably improved outcome in adult MD S[
24,
44]. However, the experience among pediatric MDS is far more limited with anecdotal report s[
22,
45]. In our study, an excellent response rate of 100% (100.0% achieved mCR) was observed using decitabine-combined MMR with a median of two cycles (range, 1-3) for pediatric advanced MDS. At the same time, three patients with RAEB achieved mCR after one or two cycles of sole decitabine. More encouragingly, the strategy of low-dose decitabine-combined-MMR use proved to be very tolerable with mild non-hematologic toxicity in the pediatric population. Considering the heterogeneity of MDS and unevenly distributed subgroups, patients with advanced MDS were further extracted and analyzed to better illustrate the effect of decitabine-combined therapy bridged allo-HSCT. As a result, 13 patients with DAC + MMR treatment showed a quite inspiring survival (84.6 ± 10.0%), and none of the 11 survivals relapsed at last follow-up. DAC + MMR appears to be a promising bridge to HSCT with its high efficiency of eliminating the excess BM blasts with low toxicity. These exciting results provided a valuable clinical experience for the use of decitabine in the pediatric population.
Preemptive treatment for the minimal residual disease (MRD) is essential for preventing or substantially delaying hematological relapse after HSCT in pediatric MDS, especially in high-risk subgroups. The discovery of genome-wide DNA hypermethylation in pediatric MDS provides a rationale for DNMT inhibitors applicatio n[
9,
17]. Low-dose decitabine could directly and irreversibly inhibit the DNA methyltransferases. More intriguing potential of decitabine among antitumoral alloimmunity and pro-apoptotic effect of tumor cells has emerged in recent year s[
24‐
26,
45]. The MMR is originated from the low-dose chemotherapy consisting of low-dose cytarabine and aclarubicin combined with G-CSF, abbreviated as “CAG”, proposed in 1995. The CAG regimen achieved certain efficacy in refractory/relapse adult MDS and AM L[
12]. Even in low/intermediate risk adult MDS and AML, the CR rates of low-dose induction therapy were significantly higher than intensive chemotherap y[
46]. However, the cardiac toxicity associated with aclarubicin mainly limited to a certain extent of the application of CAG regimen. Then, alternatives with similar therapeutic effect and mild cardiac toxicity were developed, forming different regimens of MMR. The advantages of MMR may be due to the synergy of G-CSF and low-dose chemotherapy drugs. G-CSF priming could preferentially potentiate Ara-C and anthracycline-mediated cytotoxicity on myeloid tumor cells, presumably by enhancing G0 resting tumor cells into the cell cycl e[
47]. In addition, the G-CSF combination may inhibit the self-renewal capacity of myeloid tumor cells and leukemia stem cell s[
48,
49]. It will be of great interest to investigate the underline specific mechanisms in the future. Hence, the combinatorial approach of decitabine, low dose chemotherapy drugs and G-CSF is reasonable and might be an effective strategy for pediatric MDS before transplantation.
Several limitations about our study should be considered. Firstly, the fundamental limitation is that this analysis did not include patients who received chemotherapy and/or DNA-hypomethylating therapy and did not progress to transplantation. The excellent overall responses to decitabine-concomitant treatment may not be accurately attainable for each individual among the heterogeneous MDS population. Secondly, details including chemotherapy regimens, donor types and conditioning regimens vary widely, and the robustness of the results may be impaired. Thirdly, this cohort included 28 patients with a median follow-up of 53.0 months, which is not adequate enough and may lead to a considerable bias. Finally, our analysis has the intrinsic limitation related to the retrospective nature and comparison with limited historical controls. In 2018, we had registered a multicenter study of DAC + MMR for children with MDS or AML (ChiCTR1800015872) and we are struggling for large confirmatory and prospective studies to help us to clarify whether this approach can alter the natural history of the disease. Therefore, the results in the present study must be interpreted with caution and further evidence from future prospective studies is required.
In summary, our cohort shows that probably, about 71% of the children with MDS would achieve prolonged survival with allo-HSCT. Abnormal karyotype at diagnosis, high BM blast cell percentage before transplantation and severe aGVHD may indicate undesirable outcomes. CBT is not preferred, while haploidentical HSCT might be a feasible alternative when HLA-identical HSCT is unavailable. The bridging therapy of DAC + MMR was safe and well tolerated. It appears to be more effective than AML-type chemotherapy with higher mCR rate and better survival rate in childhood MDS. Our study may provide a novel and practical bridging approach for pediatric MDS with subsequent allo-HSCT. Due to the lack of randomized controlled trials, further prospective randomized study to explicitly determine the safety and efficacy of this approach in comparison with no decitabine (AML-type chemotherapy-combined HSCT or HSCT only) are required.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.