Introduction
Allogeneic hematopoietic cell transplantation (allo-HCT) is one of the most important treatment strategies for high-risk patients with acute leukemia. Although the complications and mortality associated with transplantation have decreased in recent years, transplantation-related mortality (TRM) is still the major barrier to allo-HCT [
1‐
3]. Many studies have found that 60–80% of TRM occurs within 100 days of transplantation [
3‐
6]. Recent large-scale studies in North America and Europe reported that the TRM at 100 days significantly decreased after the year 2000 [
1,
2]. In this context, according to the Center for International Blood and Marrow Transplant Research (CIBMTR) study data, the 100-day TRM of acute myeloid leukemia patients with in first complete remission transplanted with using myeloablative conditioning (MAC) regimens decreased from 15 to 6% in matched sibling donors, and from 37 to 14% in matched unrelated donors [
1]. In the European Society for Blood and Marrow Transplantation (EBMT) study, the TRM at 100 days decreased from 21.1 to 13.6% [
2].
Some common causes of early TRM included infection, toxicity, and graft-versus-host disease (GVHD). Even within the first 100 days after allo-HCT, the cause of death varies as a result of the timing. The authors of the EBMT study reported that early mortality should be divided into the first 30 days (very early), and 30–100 days (early) after transplantation [
2]. Infections and other causes accounted for more than 80% of the deaths occurring within 30 days of transplantation. Disease recurrence and GVHD accounted for 15% of the deaths. In contrast, relapses and GVHD accounted for more than 50% of the deaths between 30 and 100 days. Among non-relapse mortality, the mortality from GVHD decreased over time. However, mortality from other causes, such as infection and organ toxicity, was not significantly reduced. Similarly, an Italian study reported that mortality from acute GVHD has decreased significantly since 2001 although the mortality from infection and multi-organ failure increased [
7].
Therefore, the objective of this study was to investigate changes in early TRM in Korea, using a large dataset from the National Health Insurance Service (NHIS) and analyzing the cumulative incidence rates (CIRs) of early TRM at 50 and 100 days after allo-HCT in patients with acute leukemia. We also investigated acute leukemia and the causes of early mortality and the associated risk factors associated with early TRM.
Methods
Data collection
This study was a nationwide, population-level, historical cohort study of patients with acute leukemia who underwent allo-HCT. This data was obtained from the claims database of the NHIS of the Republic of Korea. South Korea has a universal health coverage system provided by the central government, which has been unified since 2000. The NHIS provides health insurance to more than 99% of the population. Accordingly, the NHIS has a comprehensive health database for diagnoses, treatments, procedures, and prescriptions. They provide these extensive data for use in research after the approval process. This study also obtained death related data, including the cause of death from Statistics Korea, which has a comprehensive database in connection with the NHIS. In South Korea, death registration is usually completed and confirmed by a physician. The institutional review boards of Kosin University Gospel Hospital approved this study and granted a waiver of informed consent from the study participants owing to the nature of the data from which private information was deleted. All methods were carried out in accordance with relevant guidelines and regulations along with the approval.
Study populations
We selected patients diagnosed with acute leukemia who received allo-HCT from 2003 to 2015. This study includes data on both adults and pediatric patients. Transplant data registered with the NHIS were from patients who reached complete remission prior to transplantation. The recipients and donors were typed at the allelic level for HLA-A, HLA-B, HLA-C, and HLA-DRB1 including those with fully matched or single-HLA locus mismatched transplants.
Treatment and procedures
The conditioning intensity was defined as MAC when the total body irradiation (TBI) was administered for 4 days or longer, or when busulfan was administered for 3 days or longer. In contrast, the cases in which TBI was administered for less than 4 days or busulfan for less than 3 days were classified as reduced-intensity conditioning (RIC). Rabbit anti-thymocyte globulin (ATG; Sanofi-Aventis, Cambridge, MA) was administered to patients at various dosages to prevent GVHD. ATG was given in equally divided doses for 2 or 3 days from day − 3. All patients received calcineurin inhibitors, including cyclosporine or tacrolimus, with or without short-term methotrexate as immunosuppressants to prevent GVHD. Prophylaxis against infections included low-dose acyclovir, trimethoprim−sulfamethoxazole, antifungal agents (such as fluconazole), antibiotics (such as levofloxacin), and preemptive therapy with ganciclovir for patients with cytomegalovirus infection (on the basis of antigen or DNA testing). Half of the patients received ursodiol as prophylaxis against cholestasis.
Statistical methods
The objectives of this study were to determine the CIRs of early TRM at 50 and 100 days after transplantation and to identify the causes of death and risk factors for early TRM. The CIR of early TRM was reported at a specific time after the transplant (day 50 and 100) in a landmark approach. Patients who underwent two or more transplants in that period were analyzed for the last transplant. The probabilities of mortality were estimated using cumulative incidence curves. We used the chi-square test for categorical data and independent t-test for continuous data. The causes of death were reported within 50 days of allo-HCT. We used maximally selected log-rank statistics in the maxstat function of the R software (version 3.3.2) to identify the optimal threshold to assess the survival outcomes for age and time from diagnosis to transplantation. We selected the optimal age and time from diagnosis cut-offs to be 40 years and 9 months, respectively. The probability of overall survival was estimated by the Kaplan−Meier method. Logistic regression was used for multivariate analysis. The statistical analysis was performed using the R statistical software (version 3.4.4; R Foundation for Statistical Computing) and SAS statistical analysis software (version 9.4; SAS Institute Inc., Cary, NC, USA). P-values < 0.05 with 2 sided test were considered statistically significant.
Discussion
Total hematopoietic cell transplantation conducted in South Korea doubled over 10 years from 1139 cases in 2005 to 2286 cases in 2015 [
8]. During our study period, the frequency of allo-HCT in acute leukemia also increased. In addition, the age of patients, the use of peripheral blood, RIC regimen, and the experience of iron chelation significantly increased. Globally, with advances in supportive care and RIC regimen, transplantation in elderly and high-risk patients has also increased [
1,
2,
7,
9,
10]. In North America and Europe, the average age of transplantation increased from 33 to 40 years since 2000 [
1,
2]. According to the CIBMTR data from North America, the proportion of patients over 60 years old increased from 1% in 1994–1995 to 10% in 2004–2005 [
1]. In addition, transplantations in patients with unrelated or mismatched donors, high risk disease status, and poor performance status increased [
1,
7,
9,
11].
The CIRs of early TRM at 50 and 100 days for patients with acute leukemia between 2003 and 2015 were 2.9 and 8.3%, respectively. Other studies have reported the TRM at 100 days of 5–20%, and our results were similar [
1,
2,
11]. In addition, many studies reported a significant decrease in the mortality of transplantation over time [
1,
2,
7,
9,
10]. These studies explained these changes as a result of the use of less toxic conditioning, accurate HLA matching, advances in the prevention and treatment of GVHD, and improved engraftment with increased peripheral blood use [
1,
7,
9,
10]. In our study, there was no significant decrease in early TRM over time. During this study period, the number of patients receiving iron chelation therapy increased, while the number of elderly patients and patients who had previous transplantation also increased. On the other hand, the use of bone marrow decreased. In addition, although we have not investigated, unrelated transplantation would have increased as in other studies [
1,
9]. For this reason, we speculated that there was no significant improvement in early TRM in our study.
This study showed that the most common causes of death within 50 days of transplant were infection (pneumonia, sepsis) and organ failure. In previous studies, the most common causes of early TRM were infection (pneumonia), organ failure, GVHD, and relapse [
2,
7]. However, similar to our findings, other groups have found that infection or organ failure were related to death at very early periods after transplantation [
2,
12].
In this study, there were significantly higher CIRs of early TRM in the following settings: older age, a long time from diagnosis to transplantation, previous transplantations, the use of cord blood as a graft source, and the absence of iron chelation therapy before transplantation. Many studies have evaluated the risk factors related to TRM, including age, disease status, donor matching, stem cell source, and interval between transplants [
3,
5‐
7,
9,
11‐
15]. Old age was an important risk factor for early TRM [
1,
2,
6,
11]. In our study, age over 40 years old was a factor affecting early mortality. However, TRM was not higher in patients over 60 years old than it was in those under 60.
In addition, cord blood transplantation was associated with higher early TRM. Patients receiving cord blood usually had longer neutrophil recovery time than did those receiving bone marrow or peripheral blood [
16]. The delayed engraftment can result in high early mortality caused by neutropenic fever or sepsis. Patients who underwent previous transplantations presented with relapse or refractory disease status after the first transplantation. Patients who underwent second transplantation showed non-relapse mortality greater than 30% [
17,
18]. Patients with a high risk of relapse or refractory disease who underwent previous transplantations had a high incidence of TRM [
12].
Another significant factor for TRM was iron chelation treatment. Iron overload was already known as a major adverse prognostic factor in transplantation of benign hematologic diseases such as thalassemia [
19]. In addition, it has been reported to be associated with low survival and high TRM in allo-HCT of hematologic malignancies [
20‐
25]. Pullarkat et al. found that early mortality at 100 days and the risk of death (due to acute GVHD and infection) increased when pre-transplantation ferritin levels were higher than 1000 ng/mL [
23]. Deaths from iron overload were caused by organ toxicity and liver toxicity, such as venous occlusive disease [
24,
26]. Iron chelation has been found to reduce mortality in patients who were at risk of high mortality due to iron overload [
27,
28]. Sivgin et al. reported that peri-transplant mortality at 100 days after allo-HCT was 18.9% in patients who did not receive iron chelation and 2.3% in patients who received iron chelation therapy [
27].
In our study, a total of 14.7% of patients received iron chelating agents before transplantation. The CIR of early TRM was significantly lower in patients who received iron chelating agents than it was in those who did not receive this therapy. We did not analyze the pre-transplant ferritin levels or the duration of iron chelation in this study. Although the average number of RBC transfusions was significantly higher in patients who underwent iron chelation than it was in those who did not, the mortality rate was low. Some authors have reported that severe iron overload itself was detrimental, but also that toxic non-transferrin-bound iron caused by conditioning was associated with tissue damage [
20,
29]. Under this background, Armand et al. administered deferoxamine for 2 weeks before transplantation to 5 patients with median ferritin level of 3746 ng/mL [
30]. Veno-occlusive disease did not occur in all of them and all survived until 22 months. In general, the iron overload rate of patients before receiving allo-HCT was as high as 30–70% [
25,
31]. Although the use of iron chelation has increased, more active treatment for iron overload is needed.
This study had several limitations. First, it was limited to patients who were registered with the NHIS in Korea. In addition, because of the limitation of big data, we were not able to analyze disease status, donor type, recurrence, or detailed clinical findings. However, this study was a meaningful retrospective study that was based on large-scale transplant data conducted in Korea over 14 years. In conclusion, the CIRs of early TRM at 50 and 100 days were similar to those reported in previous studies (2.9 and 8.3%). The most common causes of death were infection and organ failure. The highest rates of early TRM were found in patients who were older, had a long period to transplantation, underwent previous transplantations, and received cord blood as the graft source. Patients who received iron chelation therapy before transplantation had a low incidence rate of early TRM.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.