Objectives: Antithymocyte globulin is extensively used for prophylaxis of graft-versus-host disease in patients undergoing haploidentical hematopoietic stem cell transplantation. However, different doses of antithymocyte globulin are administered in clinical practice. This study aimed to identify the optimal dose of antithymocyte globulin (thymoglobulin) in haploidentical hematopoietic stem cell transplantation.
Materials and Methods: We retrospectively analyzed the effects of 10 mg/kg (2.5 mg/kg on days -5 to -2) versus 7.5 mg/kg thymoglobulin (2.5 mg/kg on days -4 to -2) on patients receiving haploidentical hematopoietic stem cell transplantation with myeloablative conditioning.
Results: We observed significant differences between the 2 treatment groups with regard to cumulative incidence of grade II to IV acute graft-versus-host disease (15.3% vs 14.6%; P = .93) and 3-year chronic graft-versus-host disease (12.1% vs 14.3%; P = .77). The probabilities of 3-year overall survival (68.9% vs 73.5%; P = .98) and graft-versus-host disease-free/relapse-free survival (66.7% vs 53.1%; P = .14) were comparable between the 2 groups. However, there was a trend for lower cumulative incidence of hemorrhagic cystitis in the 7.5 mg/kg treatment group compared with the 10 mg/kg treatment group (40.7% vs 24.4%; P = .07).
Conclusions: For patients who received a reduced dose of antithymocyte globulin (7.5 vs 10 mg/kg), there was no impaired effect on prophylaxis of graft-versus-host disease, with a trend of reduced incidence of hemorrhagic cystitis. Further studies of the 7.5 mg/kg dose of antithymocyte globulin are warranted for patients receiving haploidentical hematopoietic stem cell transplantation.
Key words : Anti-T-lymphocyte globulin, Conditioning regimen, Graft-versus-host disease, Thymoglobulin
Hematopoietic stem cell transplantation (HSCT) remains a curative treatment for many types of hematologic malignancies. For patients who are in urgent need of transplant but in whom identification of a human leukocyte antigen (HLA)-matched sibling or unrelated donor is not possible, haploidentical HSCT (haplo-HSCT) is an important option with the advantage of widespread availability and quick initiation. Nevertheless, graft-versus-host disease (GVHD) is still the main problem in the application of haplo-HSCT, with a significant impact on mortality.1
Antithymocyte globulin (ATG), an important in vivo T-cell depletion strategy, has been used for GVHD prophylaxis for decades and has shown significant clinical efficacy.2-8 The main concern with clinical administration of ATG is that lower ATG exposure may decrease the effect of GVHD prevention, whereas a higher dose may increase the risk of side effects, such as delayed immune reconstitution and reduced graft-versus-tumor effects.9 The use of ATG or a higher dose of ATG have been found to be associated with infection,3-5 relapse,2,3 posttransplant lymphoproliferative disorder (PTLD),5 and delayed engraftment.4-6 Therefore, it is essential to identify the optimal dose of ATG with respect to maximized efficacy and minimized toxicities.
Thymoglobulin, one of the most extensively used ATG, is a purified polyclonal immunoglobulin G derived from rabbit immunized with human thymocytes. Both 10 mg/kg and 7.5 mg/kg thymoglobulin have been widely used in haplo-HSCT, and both have led to good clinical results.10,11 In this study, we retrospectively compared the clinical effects of these 2 doses of thymoglobulin (7.5 vs 10 mg/kg) in patients receiving haplo-HSCT with myeloablative conditioning.
Materials and Methods
This was a single-center, retrospective analysis of patients who developed hematological malignancies and underwent haplo-HSCT with granulocyte colony-stimulating factor primed peripheral blood stem cells after myeloablative conditioning. Patients with aplastic anemia, those who used other ATG and not thymoglobulin (Genzyme Polyclonals), and those who were undergoing second transplants were excluded. This study was in accordance with the Helsinki Declaration of 1975 and approved by the institutional ethics board of Shandong Provincial Hospital (Jinan, China).
This study included 100 consecutive patients, and all of them received 10 mg/kg ATG (ATG-10 group) or 7.5 mg/kg ATG (ATG-7.5 group), with 2.5 mg/kg/day administered on days -5 to -2 or on days -4 to -2, respectively. The different doses of ATG administered during the 6-year inclusion period reflected the different views of institutional physicians and changes in institutional practice because no solid evidence had demonstrated a preferred dose of thymoglobulin in a haplo-HSCT setting. Both groups were given myeloablative conditioning and standard GVHD prophylaxis, including cyclosporin, mycophenolate mofetil, and short-term methotrexate. Treatment of GVHD was according to standard protocols and the discretion of physicians for varying disease conditions. Prevention and treatment of infections, including Epstein-Barr virus (EBV) and cytomegalovirus (CMV), were done in accordance with institutional practices.
Outcomes and definitions
Acute GVHD was accessed according to the Glucksberg scale,12 and chronic GVHD was evaluated according to National Institutes of Health criteria.13 Reactivation of EBV and CMV was defined as EBV DNA load above 5000 copies/mL and CMV DNA load above 500 copies/mL in peripheral blood mononuclear cells (PBMC), respectively. The time of neutrophil and platelet engraftment was the first of 3 consecutive days that absolute neutrophil count exceeded 0.5 × 109/L and platelet count exceeded 20 × 109/L without transfusion, respectively. Overall survival (OS) was calculated as the time from transplant to death from any cause. Relapse-free survival (RFS) was calculated as the time from transplant to death or relapse. Graft-versus-host disease-free/relapse-free survival (GRFS) was defined as the time from transplant to the occurrence of grade III to IV acute GVHD or systemic therapy-requiring chronic GVHD or relapse.
Patient and transplant characteristics were compared using the chi-square test for categorical variables and the Mann-Whitney test for continuous variables. The Kaplan-Meier method was used to estimate the probability of survival (OS, RFS, and GRFS). To calculate the cumulative incidences of relapse, GVHD, CMV reactivation, EBV reactivation, and hemorrhagic cystitis, death from other causes was treated as competing risk for Gray test. Factors included in univariate analysis were patient age at HSCT (<40 vs ≥40 years), disease risk index, response status before HSCT, donor-recipient sex match, donor-recipient ABO match, infused CD34+ cells (<3 × 106/kg vs ≥3 × 106/kg), and acute GVHD (none or grade I acute GVHD vs grade II-IV acute GVHD). Multivariate analysis was evaluated using Cox proportional hazards regression analysis, and criterium for exclusion of independent variables was set at P > .10. P < .05 based on 2-sided hypothesis test was considered statistically significant. Statistical analyses were primarily performed with SPSS version 23, and R version 3.6.2. was used for competing risk analysis.
Our study included 59 and 41 patients given ATG at a total dose of 10 mg/kg and 7.5 mg/kg, respectively. Median follow-up was 366 days (range, 101-1686 days) and 540 days (range, 59-1894 days) in the 2 groups, respectively (P = .34). Only 1 patient in the ATG-7.5 group had less than 100 days of follow-up because of death. Disease risk index was used to stratify patients14 but was not suitable to evaluate the disease risk for 2 patients with chronic myelomonocytic leukemia in the ATG-7.5 group, 2 patients with mixed-phenotype acute leukemia, and 1 patient with chronic myelomonocytic leukemia in the ATG-10 group. One patient in the ATG-7.5 group had EBV reactivation before HSCT. In the ATG-10 group, 6 patients had EBV reactivation and 1 patient had CMV reactivation before HSCT. All donors did not have CMV and EBV reactivation before collection of peripheral blood stem cells. The 2 groups were equivalent with regard to patient, disease, and transplant characteristics, which are detailed in Table 1.
Acute and chronic graft-versus-host disease
Table 2 summarizes clinical outcome assessments. At 100 days after transplant, the cumulative incidence of grade II to IV acute GVHD was 15.3% (95% CI, 6.0-24.5%) in the ATG-10 group and 14.6% (95% CI, 3.7-25.6%) in the ATG-7.5 group (P = .93) (Figure 1a). The 100-day cumulative incidence of grade III to IV acute GVHD was not significantly different (P = .07) in the ATG-10 group (1.7%; 95% CI, 0.1-8.0%) versus the ATG-7.5 group (9.8%; 95% CI, 3.1-21.2%). The cumulative incidence of total chronic GVHD at 3 years after HSCT was 12.1% (95% CI, 2.9%-21.4%) and 14.3% (95% CI, 2.4%-26.2%) in the ATG-10 and the ATG-7.5 groups (P = .77), respectively (Figure 1b). Among patients with chronic GVHD, 2 of 5 in the ATG-10 group and 3 of 5 in the ATG-7.5 group had extensive chronic GVHD (P > .999).
Relapse, survival, and cause of death
The 3-year cumulative incidence of relapse was 18.4% (95% CI, 6.3-30.4%) in the ATG-10 group and 31.0% (95% CI, 15.2-46.7%) in the ATG-7.5 group (P = .22). The dose of ATG was added to the multivariate model, despite it not being significant to predict relapse in the univariate model. In the multivariate analysis, the dose of ATG did not influence the risk of relapse, whereas disease status of not in complete remission (NCR) was associated with high risk of relapse (hazard ratio [HR] = 3.04; 95% CI, 1.10-8.45; P = .03; Table 3).
The probability of 3-year OS was comparable between the 2 groups, which was 68.9% (95% CI, 52.2-85.6%) in the ATG-10 group and 73.5% (95% CI, 58.4-88.6%) in the ATG-7.5 group (P = .98) (Figure 2a). The probability of 3-year RFS (71.1% [95% CI, 57.2-85.0%] vs 69.0% [95% CI, 53.5-84.5%]; P = .88, Figure 2b) and GRFS (66.7% [95% CI, 52.0-81.4%] vs 53.1% [95% CI, 37.0-69.2%]; P = .14, Figure 2c) did not differ between the ATG-10 and the ATG-7.5 groups. For patients with status of complete remission before HSCT, the probability of 3-year GRFS was 66.2% (95% CI, 50.3-82.1%) and 58.4% (95% CI, 40.0-76.8%) in the ATG-10 and ATG-7.5 groups, respectively.
During the follow-up period, 12 of 59 patients (20.3%) and 9 of 41 patients (22.0%) patients died in the ATG-10 and the ATG-7.5 groups, respectively (P = .85). Among patients who died, the incidence of nonrelapse-related mortality was higher in the ATG-10 group, although no significant difference was observed (50% vs 11.1%; P = .16). Table 2 shows the causes of death among the study patients.
Epstein-Barr virus and cytomegalovirus reactivation and hemorrhagic cystitis
The cumulative incidence of EBV reactivation within 300 days in the ATG-10 group (91.5%; 95% CI, 84.0-99.0%) was similar to that shown in the ATG-7.5 group (90.2%; 95% CI, 80.6-99.9%) (P = .82; Figure 3a). Reactivation of EBV developed at a median time of 36 days (range, 10-103 days) and 31 days (range, 9-102 days) after HSCT in the ATG-10 and the ATG-7.5 groups, respectively (P = .24). Three patients (5.1%) in the ATG-10 group developed EBV-related PTLD at 39, 43, and 80 days after HSCT; in contrast, no patients in the ATG-7.5 group developed PTLD (P = .14). Similarly, there was no significant difference in the 300-day cumulative incidence of CMV reactivation between the 2 groups (91.9% [95% CI, 84.3-99.5%] vs 82.9% [95% CI, 70.9-94.9%]; P = .17) (Figure 3b). The median time to CMV reactivation was 31.5 days (range, 11-94 days) and 33 days (range, 12-89 days) after HSCT in the ATG-10 and the ATG-7.5 groups, respectively (P = .77). None of the patients in the 2 groups developed CMV disease.
There was a trend toward lower cumulative incidence of 100-day hemorrhagic cystitis in the ATG-7.5 group (24.4%; 95% CI, 11.1-37.7%) compared with the ATG-10 group (40.7%; 95% CI, 28.0-53.4%; P = .07) (Figure 3c). In multivariate analysis, the dose of ATG did not predict hemorrhagic cystitis(HR = 0.54; 95% CI, 0.27-1.11; P = .09); however, grade II to IV acute GVHD was associated with hemorrhagic cystitis (HR = 2.32; 95% CI, 1.08-4.97; P = .03; Table 3). Median time from transplant to occurrence of hemorrhagic cystitis was 34 days (range, 10-64 days) in the ATG-10 group and 40 days (range, 9-85 days) in the ATG-7.5 group (P = .09).
A total of 100 patients (100%) achieved sustained myeloid engraftment. The median time to achieve myeloid engraftment was 12 days (range, 6-20 days) and 12 days (range, 10-21 days) in the ATG-10 and the ATG-7.5 groups, respectively (P = .91). During the follow-up period, 1 patient in the ATG-7.5 group who was in the state of relapse before transplant did not achieve platelet engraftment and relapsed at 27 days posttransplant. Except for this patient, median time to platelet engraftment was 13 days (range, 11-56 days) and 12 days (range, 8-39 days) in the ATG-10 and the ATG-7.5 groups, respectively (P = .08). Twenty-seven patients (45.8%) in the ATG-10 group and 23 patients (56.1%) in the ATG-7.5 group achieved platelet count greater than 100 × 109/L at 30 days posttransplant (P = .42).
Haploidentical HSCT is a viable approach since almost every patient has a haplotype-sharing relative. The use of this approach was once limited by severe GVHD, which is one of the main life-threatening complications.1 In recent years, the total number of haplo-HSCT cases has increased rapidly15 because of the development of several novel prophylaxis approaches, including granulocyte colony-stimulating factor-primed allografts combined with ATG (also termed as the “Beijing Protocol”).
To our knowledge, thymoglobulin at a total dose of 10 mg/kg was first used in haplo-HSCT.10 In 2 subsequent prospective studies that enrolled patients with acute myeloid leukemia and Philadelphia-negative high-risk acute lymphocytic leukemia, haplo-HSCT with the use of 10 mg/kg thymoglobulin was shown to achieve comparable outcomes to HLA-matched sibling donor transplant regarding 3-year OS, disease-free survival, and relapse and nonrelapse-related mortality.16,17
Thymoglobulin at a total dose of 6 to 15 mg/kg has been used as a part of conditioning regimens in haplo-HSCT by different institutions5,7,11,16-23; however, only a handful of studies have compared different doses of ATG in haplo-HSCT.5,7,23 The Peking University group compared the effects of 10 mg/kg versus 6 mg/kg thymoglobulin in a prospective randomized trial of patients undergoing haplo-HSCT with myeloablative conditioning.5,7 In the trial, patients who received 10 mg/kg had significantly lower cumulative incidences of grades II to IV acute GVHD (25% vs 41.9%; P = .005), grade III to IV acute GVHD (4.5% vs 16.1%; P = .005), and 5-year chronic GVHD (56.3% vs 75.0%; P = .007).5,7 However, thymoglobulin at a dose of 10 mg/kg led to significantly delayed early immune reconstitution and increased incidence of EBV reactivation (25.3% vs 9.6%; P = .001).5 Moreover, the incidence of infection-related death was higher in those who received 10 mg/kg (14.3% vs 7.1%; P = .084), which may explain why the beneficial effect of 10 mg/kg thymoglobulin on the control of acute GVHD and chronic GVHD did not translate to higher OS and lower nonrelapse-related mortality.5,7
Given that haplo-HSCT with 7.5 mg/kg thymoglobulin also produced promising results, we wondered whether this dose could reduce the risk of infection without increasing GVHD and thus lead to better survival.11 In the present study, we found that 7.5 mg/kg thymoglobulin did not increase the incidence of acute GVHD and chronic GVHD compared with 10 mg/kg. Similarly, several studies that compared different doses of thymoglobulin (6 mg/kg and 7.5 mg/kg or 8 mg/kg), administrated in the reduced-intensity conditioning setting, showed consistent results.3,24,25 Another study showed no difference in the incidences of grade II to IV acute GVHD and total chronic GVHD between groups given 5 mg/kg, 7.5 mg/kg, and 10 mg/kg thymoglobulin who were undergoing unrelated donor HSCT.26 Given the different transplantation settings, the retrospective nature of these studies, and the limited follow-up time for chronic GVHD, the results need to be interpreted with caution.
The administration of ATG may increase the risk of relapse, as GVHD is linked to graft-versus-tumor effects.27 Remberger and colleagues found a higher incidence of relapse after reduced-intensity conditioning HSCT with unrelated donors in patients who received 8 mg/kg thymoglobulin versus patients who received 6 mg/kg thymoglobulin (41% vs 19%; P = .04); however, these results were not validated in the multivariate analysis.3 In contrast, several studies, including a prospective randomized trial, reported that patients who received higher doses of thymoglobulin did not have increased cumulative incidences of relapse.5,7,24-26 In the present study, the dose of ATG was not associated with relapse, but status of patients before HSCT was an independent influencing factor for relapse. The relatively higher proportion of NCR patients in the ATG-7.5 group may explain the trend of higher incidence of relapse. In addition, we observed a lower probability of GRFS in the ATG-7.5 group, although not significant, which may be attributed to the slightly higher rate of relapse in the ATG-7.5 group. To rule out the effect of NCR status on GRFS, we conducted a subgroup analysis of patients in complete remission. The results showed that the GRFS gap between the 2 groups narrowed, indicating that the low GRFS in the ATG-7.5 group may be due to the higher proportion of NCR patients.
Antithymocyte globulin has diverse effects on immunomodulation, not only depletion of T-cell but also induction of B-cell apoptosis and interference with antigen-presenting cell functions.28 Higher doses of thymoglobulin have been shown to result in significantly higher incidences of infectious complications, such as bacterial infection and aspergillus infection.25,26 In a retrospective study, as the dose of ATG increased (5, 7.5, and 10 mg/kg), the rate of infection-related mortality also increased significantly (3.7%, 19.0%, and 26.7%; P = .02).26
High-dose ATG is an important risk factor for viral reactivation and viral infection.29,30 A prospective investigation found that 10 mg/kg ATG could lead to delayed early immune reconstitution and result in higher 1-year cumulative incidence of EBV reactivation than 6 mg/kg ATG.5 In contrast to some reports that indicated that higher doses of thymoglobulin could increase EBV and CMV reactivation,3,5,23,25 in our study, the incidences of EBV and CMV reactivation were comparable between our 2 treatment groups, which was consistent to several previous studies.24,26 The differences in definition and sample type for detection may account for such different results. As an example, a study comparing 7.5 mg/kg and 10 mg/kg ATG set the threshold for EBV DNAemia at 500 copies/mL in plasma and found that the 1-year incidence of EBV DNAemia was 20.7% and 40.0%, respectively (P < .001).23 In our study, however, we defined EBV reactivation as EBV DNA load in PBMC greater than 5000 copies/mL, EBV DNA load in PBMC has been shown to be higher than in plasma31 and correlation coefficient is low between PBMC and plasma.32 One study showed that 45% of HSCT patients with positive EBV DNA in PBMC had negative test results in plasma.32 Hence, the incidences of EBV reactivation in this study were high, which also reduced the ability to detect meaningful differences between the 2 groups.
Increased doses of thymoglobulin were reported to result in higher incidences of hemorrhagic cystitis, and high-dose ATG has been shown to be a risk factor of hemorrhagic cystitis.24,25,33 Similarly, in our study, the ATG-10 group had a trend of higher incidence of hemorrhagic cystitis, but the multivariate analysis failed to demonstrate that dose of ATG was associated with hemorrhagic cystitis.
This study is retrospective in nature and limited by the small number of patients. Several problems need to be clarified in future studies, particularly in randomized controlled trials. These first include identifying the best dose of ATG based on different kinds of conditioning regimens, the type of donors, and the source of stem cells, since the effect of ATG at the same dose may differ according to the HSCT settings.34 Second, development of methods for personalized precision of ATG is needed. Kennedy and colleagues have taken a step forward by presenting an interaction between recipient peripheral blood absolute lymphocyte count and ATG dosing.26 In patients with low absolute lymphocyte count on the first day of ATG administration, a lower dose of ATG was suggested to decrease the risk of death.26
The application of ATG has greatly reduced the incidence of GVHD and promoted the development of haplo-HSCT. Nevertheless, concerns remain about the side effects of ATG as the graft-versus-leukemia effect is related to GVHD and ATG may inhibit immune reconstitution.27,35,36 Although many studies have explored the optimal dose of ATG, findings are still not completely clear.5,7,23-26,29,30 So far, the novel dose of ATG is different in different transplantation settings, which vary with different conditioning regimens, type of graft, and type of donor. In this study, we focused on peripheral blood haplo-HSCT with myeloablative conditioning regimens and compared the effects of 10 mg/kg versus 7.5 mg/kg ATG. Our comparison of the occurrence of grade III to IV acute GVHD between the 2 groups was limited by the small sample size and the low incidence of severe acute GVHD, and further validation in a large-sample prospective clinical trial is suggested. We indicated that 7.5 mg/kg ATG seems to be better because it did not increase the incidence of grade II to IV acute GVHD and chronic GVHD and tended to reduce hemorrhagic cystitis. These results provide a theoretical basis for clinical application and further scientific research.
Reducing the dose of ATG from 10 mg/kg to 7.5 mg/kg did not compromise GVHD prevention, impair survival, or increase the risk of relapse. Our study supported the use of 7.5 mg/kg ATG in haplo-HSCT with myeloablative conditioning.
DOI : 10.6002/ect.2021.0003
From the 1Department of Hematology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan, Shandong; the 2Department of Hematology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong; the 3State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Tianjin; and the 4School of Medicine, Shandong University, Jinan, Shandong, China
Acknowledgements: This study was supported by the National Natural Science Foundation (No. 81270598, No. 81473486, No. 81770210, and No. 81700159), the Key Research and Development Program of Shandong Province (No. 2018CXGC1213), the Technology Development Projects of Shandong Province (No. 2017GSF18189, No. 2016GSF201029), the Technology Projects of Jinan (No. 201704092), the Medical and Health Technology Innovation Program of Jinan (No. 202019044), and the Taishan Scholar Foundation of Shandong Province, Shandong Provincial Engineering Research Center of Lymphoma, Key Laboratory for Kidney Regeneration of Shandong Province. The authors have no conflicts of interest to declare.
Corresponding author: Xiaosheng Fang, Department of Hematology, Shandong Provincial Hospital, Cheeloo College of Medicine; Department of Hematology, Shandong Provincial Hospital Affiliated to Shandong First Medical University. No.324, Jingwu Road, Jinan, Shandong, 250021, China
Phone: +86 531 68778331
Table 1. Patient and Donor Characteristics
Table 2. Assessment of Clinical Outcomes
Table 3. Multivariate Analysis for Relapse and Hemorrhagic Cystitis
Figure 1. Cumulative Incidence of Graft-Versus-Host Disease
Figure 2. Probability of Overall Survival, Relapse-Free Survival, and Graft-Versus-Host Disease-Free/Relapse-Free Survival
Figure 3. Cumulative Incidences of Epstein-Barr Virus and Cytomegalovirus Reactivation and Hemorrhagic Cystitis