TO THE EDITOR: Chronic graft-versus-host disease (GVHD), one of the major hurdles in the way of successful hematopoietic cell transplantation (HCT), has increased in incidence with the widespread use of peripheral blood (PB) grafts and alternative donors, along with an increased number of older transplant recipients [1]. Although anti-thymocyte globulin (ATG) plays a protective role against GVHD across various transplant settings, including alternative donor PB HCT with reduced intensity conditioning (RIC), its optimal dosing in a specific transplant platform remains largely unknown [23].

We hypothesized that the impact of different ATG doses can depend on the disease risk index (DRI). The present study aimed to explore this hypothesis by comparing transplant outcomes between total ATG doses of 6 mg/kg and 9 mg/kg in a homogenous population stratified by DRI. These patients received PB grafts from alternative donors after a specified RIC regimen for acute leukemia or myelodysplastic syndrome (MDS).

We retrospectively identified 130 eligible patients who had undergone their first HCT between February 2008 and March 2017 at Seoul National University Hospital (SNUH) and Seoul National University Bundang Hospital (SNUBH). The donors included 10/10 human leukocyte antigen (HLA) allele-matched unrelated donors (MUDs), 7/10 or 8–9/10 partially matched unrelated donors (PUDs), and 3–4/6 or 3–7/8 or 6/10 haploidentical familial donors (HIDs), while the graft source consisted of PB stem cells only. Conditioning included the administration of intravenous busulfan at a dose of 3.2 mg/kg on day D-7 and D-6, fludarabine at 30 mg/m2 from D-7 to D-2, and rabbit ATG (Thymoglobulin) at 2.0 or 3.0 mg/kg from D-3 to D-1. Cyclosporine A or tacrolimus were additionally used with or without methotrexate. The study protocol was reviewed and approved by the Institutional Review Boards of SNUH and SNUBH.

Baseline characteristics of included patients are summarized in Supplementary Table 1. The median follow-up period for the total population was 35.00 months [95% confidence interval (CI), 30.34–39.66]. In the total population, the GVHD-free, relapse-free survival (GRFS), disease-free survival (DFS), and overall survival (OS) tended to be longer when using the 6 mg/kg dose than when 9 mg/kg was used, but without statistical significance (Fig. 1A–C). In 99 patients with low/intermediate DRI, those in the 6 mg/kg group had significantly higher DFS and OS than those in the 9 mg/kg group, while their GRFS was similar. The estimates of 2-year GRFS, DFS, and OS rates were 31% (95% CI, 23–40) vs. 25% (95% CI, 19–30; P=0.133; Fig. 1D), 65% (95% CI, 57–74) vs. 43% (95% CI, 37–49; P=0.017; Fig. 1E), and 73% (95% CI, 65–81) vs. 53% (95% CI, 46–59; P=0.018; Fig. 1F), for the 6 mg/kg group vs. 9 mg/kg group, respectively. In contrast, for 31 patients with a high or very high DRI the GRFS, DFS, and OS did not differ significantly between the different ATG doses. The estimates of 2-year GRFS, DFS, and OS rates were 20% (95% CI, 7–33) vs. 29% (95% CI, 19–39; P=0.999; Fig. 1G), 20% (95% CI, 7–33) vs. 38% (95% CI, 28–49; P=0.386; Fig. 1H), and 30% (95% CI, 16–45) vs. 38% (95% CI, 28–49; P=0.855; Fig. 1I), for the 6 mg/kg group vs. 9 mg/kg group, respectively.

Grade III–IV acute GVHD at day 100 and chronic GVHD requiring systemic therapy after 2 years of HCT were more frequently noted in the ATG 6 mg/kg group than in the 9 mg/kg group, with a cumulative incidence of 23% vs. 16% (P=0.023) and 41% vs. 21% (P=0.025), respectively. The cumulative incidence estimates of relapse and non-relapse mortality at a 2-year time point were numerically lower when an ATG dose of 6 mg/kg was used than when 9 mg/kg was used, with the cumulative incidence being 17% vs. 24% (P=0.517) and 29% vs. 34% (P=0.311), respectively, albeit without statistical significance (Supplementary Table 2).

Subsequent multivariable analyses showed that patient sex (male), a higher HCT-comorbidity index (HCT-CI), the 9 mg/kg ATG dose, and high or very high DRI were independently associated with a worse survival, with hazard ratios (HRs) of 1.85 (95% CI, 1.11–3.09; P=0.019), 1.36 (95% CI, 1.06–1.76; P=0.018), 4.14 (95% CI, 1.52–11.26; P=0.005), and 2.56 (95% CI, 1.19–5.51; P=0.016), respectively, while donor type did not harbor this association. Additionally, the interaction between ATG dose and DRI was a significant predictor of OS, with a HR of 0.26 (95% CI, 0.08–0.85; P=0.026; Table 1).

Our findings are consistent with those of previous studies. Remberger et al. reported a lower incidence of GVHD and a higher incidence of relapse with an ATG dose of 8 mg/kg compared to 6 mg/kg in RIC HCT from a PB or bone marrow (BM) graft of MUDs, in accordance with our study (Supplementary Table 2) [4]. Chang et al. [5] conducted a prospective randomized trial comparing ATG doses of 6 mg/kg and 10 mg/kg in HCT from both PB and BM grafts from HIDs. These researchers found that although a higher ATG dose was associated with better GVHD prevention, it increased the risk of infectious complications, which was similarly observed in the results of the current analysis (Supplementary Table 2); however, Chang et al. [5] only included standard-risk disease and used myeloablative conditioning. A recent Korean study also suggested that ATG doses ranging from 2.5 to 7.5 mg/kg were associated with better survival in comparison to doses ranging from 9 to 12 mg/kg in recipients of mismatched HCT for acute leukemia or MDS [6].

Interestingly, in the present study, the survival benefit of the 6 mg/kg ATG dose over the 9 mg/kg dose was apparent only in patients with low or intermediate DRI, but not in those with high or very high DRI. Indeed, RIC HCTs for hematologic malignancies are expected to exert its main therapeutic effects through graft-versus-leukemia (GVL) effects, and a close association between chronic GVHD and GVL has been noted previously [78]. Because patients with high or very high DRI are at a higher risk of relapse than those with low or intermediate DRI, an ATG dose of 6 mg/kg may have been excessive for them.

Our study has several limitations. First, the retrospective nature of this study may render this analysis hypothesis-generating at most. Second, this study did not address the interaction between ATG dose and donor types. Although donor type was not significantly associated with survival in the multivariable analysis, subtle differences may have not been detected due to the limited number of patients included in the study.

In conclusion, the present study suggests that a total ATG dose of 6 mg/kg is more suitable than that of 9 mg/kg in RIC PB HCT from alternative donors in patients with acute leukemia or MDS and low or intermediate DRI, while those with high or very high DRI may require a more cautious strategy on GVHD prophylaxis.