We evaluated here the strategies to reduce GvHD risks in allogeneic HSCT by using ATG-F at two different dose levels and in different patient populations. Our main finding was that lower dose ATG-F prolonged survival, decreased TRM and reduced the incidence of cGvHD especially in the first year after transplantation, but did not affect incidence of aGvHD. In contrast, higher dose ATG-F did not induce better survival or lower TRM, suggesting a dose-dependent effect. Relapse rates were similar in all groups. Some dose-dependent liver and renal toxicities were found following ATG-F. Engraftment was delayed and immune reconstitution as assessed by CD4+ cell counts at day 100 was lower with ATG-F at both dose levels. Consistently, the incidence of infection was higher in the ATG-F groups. Especially EBV but also CMV reactivation incidence was higher in the ATG-F groups, viral infections were more common and bacterial BSI appeared to be more frequent. No difference was found in invasive fungal infections. From our data we cannot decide if this is based on different biologic effects or mainly due to the low number of fungal infections in all three groups.

Our interpretation is that ATG-F may reduce some of the complications of allogeneic HSCT by diminishing risks of cGvHD, and thereby lead to improved results particularly when the lower dose is used. This possible benefit comes at the cost of poorer immune reconstitution, more infectious risks and some direct toxicity.

Groups differed by some variables, as disease types, donors and conditioning regimens were not equally distributed among groups. However, careful adjustment for covariates did not change the main results of this analysis. Although unrelated donor transplants had a somewhat poorer survival, donor type was not significant in our multivariate model, demonstrating that differences in outcome between unrelated and related donors were successfully reduced in recent years. No interactions were noted between lower dose ATG-F and type of donor, indicating that ATG-F effects were independent of donor type. Several studies confirm these data, showing similar outcomes regardless of donor type.18, 19, 20

We found a survival benefit with lower dose ATG-F. Previous studies, of which some are published as abstracts only, found some advantage for ATG treatment. However, with the exception of one study with matched-related HSCT only,21 these studies did not reach statistical significance.7, 8, 9, 10, 13, 22 The survival benefit shown here may be due to an improved selection of patients receiving ATG. Although the imbalance of known risk factors was adjusted for by statistical analysis, larger series of patients need to be analyzed for confirmation.

Four randomized studies showed a significant decrease in acute and cGvHD after in vivo T-cell depletion with ATG.7, 10, 12, 13 In our study, we found that lower dose ATG-F decreased the incidence of cGvHD without influencing aGvHD. Similar results have been reported by Bonifazi et al.23 Figure 3 shows a lower incidence of cGvHD, mainly in the first year post transplant in the lower dose ATG-F group. Although only few patients contribute to the slight increase in the second year, it is possible that lower dose ATG-F only delays rather than eliminates cGvHD.

Engraftment was significantly slower in both ATG-F groups, possibly related to inflammation and cytokine storm. This effect had already been reported in previous studies.5, 7, 9, 10, 22, 23

Incidence of CMV and EBV reactivations, as well as symptomatic viral infections and bacterial BSI, but not invasive fungal infections were increased especially when using higher dose ATG. This can be attributed to the immunosuppressive effect of ATG. This coincides with slower immune reconstitution with CD4+ cell counts being lower in both ATG groups at day 100. Other groups had similar findings, for example, Bacigalupo et al.,10 who found more lethal infections with ATG, but others did not find any difference in infection-related mortality.7, 13, 22 Previous studies had only analyzed fatal infections or patients with >1 any infectious episode without further differentiation. We analyzed infections as viral reactivation, symptomatic viral infection, bacterial and fungal infection within 6 months from transplantation. It appears that ATG-F use does have a dose-dependent negative impact on viral infections.

We found some hepatic and renal toxicity of ATG-F, bilirubin and creatinine levels from day −5 to day +5 being significantly higher, especially in the higher dose ATG group, and might be related to cytokine release and activation of the coagulation system. Not many studies have analyzed toxic effects of ATG, but similar findings have been reported earlier.24, 25

Our study has several limitations: The study is retrospective, controls are in part historical, although the study period is reasonably short (2009–2014). The lower dose ATG group had a shorter follow-up period than the other two groups, because ATG-F was only given in the lower dose in the later study period (Table 2). Groups differed by some baseline variables, as shown in Table 1, however, we carefully adjusted for these differences in the multivariate analysis. Finally, we cannot exclude that the lower dose ATG group showing better outcome included more low-risk patients in ways not measured by our analysis. However, the fact that lower dose ATG was given to patients with unrelated donors and to patients with related donors if they were older, lends credibility to our hypothesis as these patients are considered to be at higher risks of GvHD and other complications.

In summary, our results are consistent with previous studies, including randomized trials and observational studies, showing in the majority that ATG as part of the conditioning reduces GvHD, particularly cGvHD without increasing relapse but without a survival benefit, as reviewed by Storek et al.26 In our cohort though, there is a survival benefit in patients with unrelated donor HSCT and with related donor HSCT at high risk of GvHD receiving lower dose ATG-F with less TRM in spite of higher risks of infectious complications.

In randomized trials, higher doses of ATG have been used for unrelated donor HSCT and lower doses for related transplantation.7, 10, 12, 13, 23 It is possible that using lower doses for unrelated donor HSCT and related donor HSCT at higher risks of GvHD proves to be the best treatment option.

The recently adopted strategy of using lower doses of ATG-F in unrelated and older age-related donor HSCT appears to be associated with lower TRM and lower mortality in this rather heterogenous patient population. Results therefore require confirmation by other studies.