In this study, outcomes for 575 adult ALL patients aged 45 years who underwent first allo-SCT in CR were analyzed according to the type of conditioning regimen (myeloablative conditioning (MAC) for 369 patients vs reduced-intensity conditioning (RIC) for 206 patients). Patients in the RIC group were older (median age, 58 vs 51 years, P<0.0001). There were no statistically significant differences in 3-year OS, disease-free survival (DFS) and non-relapse mortality (NRM): 51% vs 53%, 47% vs 39% and 38% vs 36%, respectively. Multivariate analysis showed that CR2 and HLA mismatching were associated with poor OS (P=0.002 and P=0.019, respectively). HLA mismatching was associated with lower rate of relapse (P=0.016), but was associated with higher rate of NRM (P=0.001). RIC was associated with good OS and DFS in patients who received HLA-mismatch transplantation and were aged 55 years compared with MAC by multivariate analysis for each event with interaction (hazard ratio (HR) and 95% confidence interval 0.35 and 0.15–0.81, P=0.014 for OS and 0.36 and 0.16–0.81, P=0.013 for DFS). Therefore, patients 55 years of age with HLA-mismatch transplantation should be candidates for RIC rather than MAC.
Although 80–90% of patients with adult ALL achieve CR, most patients relapse and die from the disease.1 Chemotherapy has resulted in long-term leukemia-free survival in 30 to 40% of ALL patients, but much higher rates of leukemia-free survival have been obtained with conventional myeloablative conditioning (MAC) allo-SCT. Recent large-scale prospective donor vs no donor studies have revealed that outcomes of matched sibling allografts were better than those of chemotherapy.2, 3, 4, 5, 6 Moreover, allo-SCT can provide better disease-free survival (DFS) not only for ALL patients in first CR (CR1) but also for those in second CR (CR2).7, 8, 9 Most conditioning regimens have included TBI, sometimes exceeding 13 Gy for patients in CR2.10, 11 We have reported excellent outcomes of allo-SCT using a conditioning regimen with medium-dose VP-16, CY and TBI (12 Gy) for adult patients with ALL.12, 13 However, non-relapse mortality (NRM) may cause a worse overall outcome of MAC allo-SCT for elderly patients and patients with comorbidities. Therefore, allo-SCT using reduced-intensity conditioning (RIC) may provide opportunities to obtain a significant GVL effect, without the adverse effects of intense myeloablative preparative regimens.14, 15, 16, 17 Marks et al.18 reported no effect of conditioning intensity on TRM or relapse risk after RIC and MAC in 93 and 1428 Ph chromosome-negative ALL patients, respectively, in first or second CR and in patients >16 years of age who received allografts from siblings and unrelated donors. Mohty et al.19 reported no effect of conditioning intensity on leukemia-free survival after RIC and after MAC in 127 and 449 ALL patients, respectively, in first or second CR and in patients >45 years of age who received allografts from HLA-identical sibling donors and were followed up for a median period of 16 months. A Japanese nationwide survey of 77 patients with hematological malignancies (aged 25–68 years) who recieved BMT after RIC from unrelated donors showed 50% OS with a median follow-up period of 439 days.20
In the current study, outcomes for 575 adult ALL patients aged 45 years at transplantation who underwent allo-SCT in CR were analyzed according to the type of conditioning regimen (MAC for 369 patients vs RIC for 206 patients) before allo-SCT.
Materials and methods
Study design and data collection
This study was a retrospective analysis of data from a Japanese nationwide multicenter survey. Data for adult ALL patients were provided by the Adult ALL Working Group of the Japanese Society of Hematopoietic Stem Cell Transplantation (JSHCT). Outcomes of 575 adult ALL patients aged 45 years at transplantation who underwent allo-SCT in CR were analyzed according to the type of conditioning regimen (MAC vs RIC) before allo-SCT.
This study included ALL patients who received MAC or RIC allo-SCT in CR: and who (1) were aged 45 years at the time of transplantation, (2) underwent transplantation between 2000 and 2009, and (3) received an MAC regimen (n=369) as high-dose radiation and chemotherapy usually in combination with CY, or a RIC regimen (n=206) defined as the use of fludarabine with low-dose TBI (8 Gy), BU (9 mg/kg) or melphalan (140 mg/m2).21
Differences between patients, disease and transplantation-related factors according to conditioning regimens and GVHD prophylaxis are shown in Table 1. As per JSHCT centers’ practice for allo-SCT for ALL, patients were eligible to receive an MAC regimen if they were aged <55 years (n=288, 78%), and 66 patients (22%) who were aged 55 years without significant comorbidities also received an MAC regimen. In the RIC group, 190 patients (92.2%) received a RIC regimen mainly because of age (50 years), regardless of the presence or absence of significant comorbidities. Sixteen patients (7%) aged <50 years received a RIC regimen possibly as a result of the physician’s decision based on significant comorbidities or some clinical reasons.
Primary end points included OS, DFS, relapse (cumulative incidence of relapse) and NRM. Relapse was defined as clinical and hematological leukemia recurrence. NRM was defined as death during continuous CR after transplantation.
Characteristics of patients who received MAC and RIC were compared using the χ2-test for categorical variables and the t-test for continuous variables. To compare the prognosis of MAC and that of RIC, univariate survival analyses were conducted for OS, DFS, NRM, cumulative incidence of relapse, engraftment (neutrophil recovery at 100 days), acute GVHD (grades II–IV) and chronic GVHD. Survival curves of OS and DFS for each group were depicted using the Kaplan–Meier method and compared by the log-rank test. In the analysis of NMR, engraftment, cumulative incidence of relapse, acute GVHD and chronic GVHD, probabilities of the incidences were calculated using the cumulative incidence function and compared by Gray’s test to accommodate competing risks.22 To adjust the potential confounders, multivariate analyses were conducted using the Cox proportional hazards model for OS and DFS, and using the Fine-Gray proportional hazards model for cumulative incidence of relapse and NRM.23 In addition, the interaction terms between treatment (MAC vs RIC) and the above confounders were included in the multivariate model for OS and DFS. If interaction terms were statistically significant (P-value <0.05), the adjusted hazard ratios were also calculated on the basis of the multivariate model that included the interaction terms as subgroup analyses. All statistical analyses were conducted using SAS ver 9.2 (SAS Institute Inc., Cary, NC, USA) and R (www.r-project.org, last accessed April 5, 2012).
Patients and clinical characteristics
Table 1 shows clinical and biological characteristics of the 369 MAC and 206 RIC patients who received allo-SCT for ALL. Patients in the RIC group were older (median age, 58 vs 51 years, P<0.0001). Seventy-six percent of the RIC patients were aged 55 years, whereas only 22% of the MAC patients were aged 55 years. More RIC patients received related peripheral blood (24% vs 13%, P<0.002), and RIC was performed more frequently in the more recent time period (61% vs 52% during 2006–2009, P=0.035). There were no significant differences in other prognostic factors such as performance status, WBC at diagnosis, cytogenetics, disease status and HLA matching.
Hematological recovery and GVHD
Engraftment (neutrophil recovery at 100 days) occurred in 92% of the MAC patients and 93% of the RIC patients (Table 2). Acute GVHD grades II–IV occurred in 44% of the MAC patients and 42% of the RIC patients (P=0.353). Moreover, chronic GVHD at 3 years occurred in 36% of the MAC patients and 35% of the RIC patients (P=0.793). There was no statistically significant difference.
OS and DFS
OS at 3 years for MAC patients was 51% and that for RIC patients was 53% (P=0.701). DFS at 3 years for MAC patients was 47% and that for RIC patients was 39% (P=0.098). There was no statistically significant difference.
There was no statistically significant difference in relapse at 1 year between the MAC and RIC groups (14% for RIC and 12% for MAC, P=0.664). However, a larger percentage of patients relapsed at 3 years in the RIC group than in the MAC group (26% for RIC and 15% for MAC, P=0.008).
NRM and cause of death
Conditioning regimen intensity had no impact on NRM at 3 years in the MAC and RIC groups (36% for RIC and 38% for MAC, P=0.678). Causes of death are shown in Table 3. Original disease and infection were the most common causes of death, followed by GVHD. Interstitial pneumonitis was more common in the MAC group.
Multivariate analysis for each event
There were no statistically significant differences in OS, DFS, relapse and NRM between the MAC and RIC groups (Table 4). CR2 and HLA mismatching were associated with poor OS (hazard ratio (HR) 1.88, P=0.002 for CR2 vs CR1 and 1.67, P=0.019 for mismatching vs matching), and female gender was associated with good OS (HR 0.59, P=0.003 for females vs males). CR2 was associated with poor DFS (HR 1.95, P<0.001 for CR2 vs CR1), and female gender was associated with good DFS (HR 0.65, P=0.006 for females vs males). CR2 was associated with higher rate of relapse (HR 2.29, P=0.007 for CR2 vs CR1). Interestingly, HLA mismatching was associated with lower rate of relapse (HR 0.27, P=0.016 for mismatching vs matching); however, HLA mismatching was associated with higher rate of NRM (HR 2.35, P=0.001 for mismatching vs matching). Female gender was associated with lower rate of NRM (HR 0.50, P=0.001 for females vs males).
When the interaction terms for each variable and the treatment were evaluated, the interaction between age or HLA status and the treatment was statistically significant. Therefore, subgroup analyses were conducted. As shown in Figure 2, RIC was associated with good OS and DFS in patients who received HLA-mismatch transplantation and were aged 55 years or more compared with MAC by multivariate analysis for each event with interaction (HR 0.35, P=0.014 for OS and 0.36, P=0.013 for DFS). Conversely, MAC showed good OS and DFS in patients with HLA matching and who were aged <50 years (HR 3.88, P=0.003 for OS and 3.51, P=0.003 for DFS).
The role of allo-SCT in adult ALL is still controversial; however, allo-SCT is a potentially curative treatment for patients with ALL. However, the majority of older adult ALL patients are not candidates for MAC regimens. Although significant reduction of the intensity of the preparative regimen may have a negative impact on long-term leukemic control,24, 25 RIC is a reasonable preparative option for older ALL patients in order to reduce regimen-related toxicities. There is little information on RIC allo-SCT for ALL patients.14, 15, 16, 17, 18, 19, 26 It was reported in 2008 by Mohty et al.14 that 2-year OS, leukemia-free survival and NRM were 52, 18 and 18%, respectively, after RIC allo-SCT for 97 adult ALL patients. RIC allo-SCT with cord blood and RIC allo-SCT with PBSCs were both feasible for adult ALL patients.15, 16 Moreover, RIC allo-SCT was suggested to be a potential therapeutic approach for adult high-risk ALL patients in remission based on the results of a prospective phase 2 study.17 Marks et al.18 found that conditioning intensity did not affect TRM or relapse risk by multivariate analysis of a comparison of 93 Ph chromosome-negative ALL patients >16 years of age after RIC with 1482 patients who received MAC. Mohty et al.19 found by multivariate analysis that NRM was decreased in RIC recipients, whereas it was associated with higher relapse rate in 576 ALL patients (RIC for 127 and MAC for 499 patients) aged 45 years. For Ph chromosome-positive ALL patients in first remission, RIC allo-SCT with post-grafting imatinib resulted in favorable long-term survival.26 Lee et al.27 reported that the BuFlu regimen (BU plus fludarabine) is not a suitable replacement for the BuCy regimen (BU plus CY) in young adults who are eligible for MAC therapy for allo-SCT.
In this study, outcomes for 575 adult ALL patients aged 45 years at the first transplantation who underwent allo-SCT in CR were analyzed according to the type of conditioning regimen (MAC for 369 vs RIC for 206). The survival rate of RIC patients was similar to that of MAC patients, despite an older median age of RIC patients. Relapse rate at 3 years was higher in the RIC group; however, OS, DFS and NRM were similar in the two groups. We divided patients into two age groups, one group with age of <55 years and one group with age of 55 years. There were no significant differences in OS and DFS between the MAC and RIC patients in the two age groups (data not shown). We found that HLA mismatching was associated with lower rate of relapse, and it seems that allo-SCT for ALL induces a GVL effect. However, HLA mismatching was associated with higher rate of NRM. RIC was associated with good OS and DFS in patients who underwent HLA-mismatch transplantation and were aged 55 years compared with MAC by multivariate analysis for each event with interaction. Conversely, MAC resulted in good OS and DFS in patients with HLA matching and who were aged <50 years. Therefore, patients with HLA-mismatch transplantation and who are aged 55 years would be candidates for RIC rather than MAC. Female gender was associated with good OS and DFS, but donor/recipient sex mismatch did not affect survival. The reason for this is not clear, but lower rate of NRM in female patients may be associated with good survival. This study has some limitations that would influence data interpretation because the patient populations were different. More of the RIC patients received PBSCs and more received a transplantation after 2006. The reason for selecting RIC is not always apparent. Therefore, our retrospective study had these serious limitations and there is a need for prospective randomized trials. However, the results of this study suggest that RIC allo-SCT is feasible and is a potential option for ALL patients aged 45 years in CR who are not eligible for MAC allo-SCT for some reason. Moreover, RIC may be a useful preparative regimen for patients aged 55 years, especially those with HLA-mismatch donors.
This study was supported in part by a Grant-in-Aid from the Ministry of Health, Labor and Welfare of Japan and a Japanese Grant-in-Aid for Scientific Research.
About this article
JT and NM designed the study and prepared the manuscript; NN and KO performed the statistical analysis; SN, KO, ST, TE, HN, Ke M, Ko M, HS, YM, KK and RS participated in interpretation of data and approval of the final manuscript.
Reduced Intensity Conditioning Allogeneic Hematopoietic Stem Cell Transplantation for Acute Lymphoblastic Leukemia; Current Evidence, and Improving Outcomes Going Forward
Current Hematologic Malignancy Reports (2018)