Clinical Experience of Allogeneic Hematopoietic Stem Cell Transplantation in Elderly Patients Aged 60 Years and Older in South Korea

Myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) are representative myeloid malignancies. There are many treatment options for MDS and AML, from best supportive care to aggressive chemotherapy. Among them, allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative treatment strategy for patients with AML or MDS. The therapeutic effect of allogeneic HSCT is thought to be largely mediated by graft-versus-leukemic (GVL) effect.

MDS and AML are known to mainly occur in the elderly, and the median age of diagnosis for both diseases is over 60 years.1 The management of elderly patients with MDS and AML is often challenging due to their age-related non-hematologic comorbidities and increased vulnerability to therapeutic toxicities.

Conventionally, myeloablative conditioning (MAC) regimens were administered before allogeneic HSCT to induce tumor cytoreduction and ideally disease eradication, and to achieve sufficient immunosuppression to overcome host rejections of the donor stem cells.2 As MAC regimens resulted with high regimen-related toxicities in the elderly and patients with comorbidities, reduced intensity conditioning (RIC) regimens, which are designed to be less ablative and less toxic, were developed. Several studies showed that RIC regimens followed by allogeneic HSCT were associated with mixed chimerism and then full chimerism with a documented GVL effect in the setting of hematologic malignancies and graft-versus-tumor effect in the setting of solid tumors. RIC is thought to rely more on GVL effect and less on cytotoxic effect for therapeutic efficacy.3 Many studies comparing MAC with RIC in patients with AML or MDS showed that RIC is associated with increased relapse but reduced treatment-related mortality, resulting in similar overall survival (OS).

Recently, Basak, et al.4 reported a retrospective analysis of the outcomes of allogeneic HSCT in a large cohort of patients aged 65 years or older reported to the European Group for Blood and Marrow Transplantation (EBMT). Their study confirmed the feasibility of allogeneic HSCT in a large-series of elderly patients, with acceptable non-relapse mortality (NRM) and OS at 1 and 3 years, including those with very advanced age of 75 years and older.

In Asia, Aoki, et al.5 reported a large-scale, nationwide retrospective study on the outcomes of allogeneic HSCT using RIC in AML patients, including a substantial number of elderly patients aged 60 years and older. However, there are still limited reports on the outcomes of allogeneic HSCT in elderly patients aged 60 years or older, especially in Asian patients. The purpose of the present study is to share our outcomes and experiences on allogeneic HSCT in elderly patients aged 60 years and older with MDS or AML in South Korea, and to compare them with other studies. We also analyzed which pretreatment factors affect OS after allogeneic HSCT.

Allogeneic HSCT was not actively performed in elderly patients with MDS or AML as conventional myeloablative allogeneic HSCT resulted with high regimen-related toxicities in them. However, the invention of RIC and improvements in supportive care have made it possible to safely conduct allogeneic HSCT in elderly patients. Under these circumstances, allogeneic HSCT using RIC on elderly patients has been increasingly performed worldwide, and the same is true in South Korea. In this study, we analyzed the outcomes of allogeneic HSCT for 116 elderly patients aged 60 years and older with MDS or AML at two centers in South Korea, and compared them with other studies.

In the present study, neutrophil and platelet engraftment were achieved at median day +11 (IQR 10–15) and +14 (IQR 11–19), respectively. These values are similar to those generally expected when HSCT is performed using peripheral blood.9 In addition, only 9 (7.8%) patients experienced primary graft failure (data not shown), which is comparable with the result of Basak, et al.4 (7.8% vs. 6.0%). All patients who failed engraftment received peripheral blood stem cells infusion. Except for two patients who received 3.6 and 3.9×10⁶/kg of CD34⁺ cells, the remaining patients received 4×10⁶/kg or more CD34⁺ cells. Although we could not explain the exact cause of graft failure in all individual patients, it is thought that various factors may have been involved in the graft failure. A complete donor chimerism was confirmed in 56.0% patients at 3 weeks and in 54.3% of patients at 3 months after HSCT. These values are similar to those reported in previous papers on allogeneic HSCT using RIC.10, 11 These results suggest that there might be no particular problem in the engraftment in allogeneic HSCT with RIC in the elderly.

The estimated cumulative incidence of grade II–IV acute GVHD at day 100 was 13.7% in our study, which is slightly lower than the result of Basak, et al.4 (13.7% vs. 28%). This difference is thought to be related to the difference in the implementation rate of T cell depletion for GVHD prophylaxis. T cell depletion was performed in 92.2% of patients in our study, but only in 66% of patients in the study of Basak, et al.4 The estimated cumulative incidence of grade II–IV acute GVHD at day 100 observed in our study was not higher than that reported in other papers on allogeneic HSCT using RIC.12, 13 In the study by Aoki, et al.5 on Japanese patients with AML who received allogeneic HSCT with RIC, the incidence of grade II–IV acute GVHD was 37.6% in patients aged 60–64 years and 31.0% in patients aged ≥65 years, and was not significantly different among the four age groups (50–54, 55–59, 60–64, and ≥65 years).

The estimated cumulative incidence of chronic GVHD at 2 years was 38.8% in our study, which is comparable to the incidence of chronic GVHD in patients aged 60–64 years (39.6%) and in patients aged ≥65 years (38.0%) reported in the study by Aoki, et al.5 The estimated cumulative incidence of chronic GVHD at 2 years in our study was also not higher than that reported in other papers on allogeneic HSCT using RIC.12, 13 Indeed, Aoki, et al.5 reported that the incidence of chronic GVHD was not significantly different among the four age groups (50–54, 55–59, 60–64, and ≥65 years).

The NRM and OS at 1 year in our study were similar to those reported by Basak, et al.4 (20% vs. 27% for NRM and 64% vs. 56.6% for OS). In a meta-analysis performed by Rashidi, et al.14 on allogeneic HSCT in AML patients over 60 years of age, the estimated NRM was 26% at 1 year and 29% at 2 years, and estimated OS was 58% at 1 year and 45% at 2 years. In the study of Pohlen, et al.15 on patients aged ≥60 years with AML and MDS who received allogeneic HSCT, NRM at 1 year was 37% and OS at 3 years was 35%. In the study of Aoki, et al.,5 the 3-year cumulative incidence of NRM was 29.2% in patients aged 60–64 years and 27.6% in patients aged ≥65 years, and was not significantly different among the four age groups (50–54, 55–59, 60–64, and ≥ 65 years).

Subgroup analysis in our study showed that male sex and HCT-CT ≥3 were independent risk factors for poor OS, and patients aged ≥65 years did not show significantly worse OS compared to those aged 60–64 years. The differences in OS according to sex and higher HCI-CI score was thought to be due to differences in the incidence of NRM, as there was no significant difference in the incidence of relapse according to sex and HCI-CI score. Pohlen, et al.15 reported that there was no significant difference in OS between patients aged 60–64 years and ≥65 years. Interestingly, in this study, HCT-CI ≥3 was predictive for worse OS only in patients aged ≥65 years, but not in patients aged 60–64 years. Aoki, et al.5 reported that OS was not significantly different among four age groups (50–54, 55–59, 60–64, and ≥65 years). Aoki, et al.5 also reported that HCT-CI ≥3 was an independent prognostic factor for poor OS. Meanwhile, Basak, et al.4 reported that patients aged ≥70 years showed slightly, but significantly poor OS compared to patients aged 65–69 years.

At present, it has not been clearly established whether male sex is a risk factor for post-HSCT mortality. However, some studies have reported that male recipients have poor post-transplantation survival compared to female recipients regardless of the donor’s sex. Representatively, Kim, et al.16 reported that male recipients had more chronic GVHD when they received stem cells from female donors and had more relapse when they received it from male donors. In our study, men tended to have more MDS diagnoses than women, but the difference was not statistically significant (male, 35/76, 46.1% vs. female, 11/40, 27.5%, p=0.052, data not shown). There was also no significant difference between male and female in other factors that could affect survival, such as age, high risk based on HCT-CI, 2 or higher grade acute GVHD, or chronic GVHD (data not shown). Based on our results, we could not explain why men had a higher mortality rate compared to female.

Given the expected NRM and OS of the elderly patients receiving allogeneic HSCT for MDS and AML, how should allogeneic HSCT be applied to which patients? In MDS, allogeneic HSCT is generally considered for patients with higher risk disease or those who failed to hypomethylating agents. As patients with higher risk MDS generally have a shorter life expectancy, allogeneic HSCT should be considered. Meanwhile, a cautious approach to transplantation is required in patients who failed to hypomethylating agents. In the study by Jabbour, et al.,17 the 1-year OS of patients who had low-risk and intermediate-1-risk MDS at the time of HMA failure was 90% and 77%, respectively. In these patients, especially the elderly, allogeneic HSCT may not be beneficial in that the expected NRM at 1 year is about 20%. In AML, allogeneic HSCT is generally considered in patients with unfavorable prognostic disease or relapse/refractory disease, as they generally have a shorter life expectancy. However, not all of these patients, especially the elderly with comorbidities, are expected to benefit from allogeneic HSCT. In our study, the expected NRM and OS at 1 year after allogeneic HSCT in patients aged ≥60 years with HCT-CI ≥3 were 35% and 47%, respectively. Based on these results, a cautious approach to transplantation in elderly patients with HCT-CI ≥3 is required. In patients aged ≥70 years, allogeneic HSCT should be determined in consideration of the patient’s medical conditions, such as comorbidities. Indeed, Lachowiez, et al.18 reported that allogeneic HSCT was beneficial in terms of OS in selected patients aged ≥70 years with high-risk MDS and AML. Meanwhile, HSCT using MAC should be considered first for patients who are in good general condition, even if they are aged over 60 years. Scott, et al.19 reported that OS was significantly better in patients with AML and MDS who received MAC compared to those who received RIC due to the significantly higher earlier relapse rates in patients who received RIC. In this paper, the 4-year relapse rate was 19.8% (95% CI, 12.7–27.9) for MAC versus 60.7% (95% CI, 51.2–69.8) for RIC. An analysis of our data on patients who received HSCT for higher risk MDS and AML showed that the relapse rate in patients aged under 60 years who received HSCT with MA was 21.3% (data not shown), which was slightly lower than the 28.4% in patients aged 60 years and older who received HSCT identified in this study.

The present study showed that allogeneic HSCT in elderly adults aged 60 years and older can be performed with successful engraftment, and acceptable NRM and OS are expected given the generally known survival of patients with higher risk MDS and poor risk AML. With careful patient selection and appropriate use of RIC, it is thought that allogeneic HSCT can be actively applied to elderly patients.