Kidney transplantation remains to be the gold standard treatment of end-stage renal disease (ESRD) ensuring significant improvements in patient survival and quality of life (1, 2). Induction immunosuppressive treatment after kidney transplantation, including T-cell– or non-T-cell–depleting therapy, reduces the incidence and severity of acute rejection, delays the initiation of calcineurin inhibitors, and/or helps keep maintenance corticosteroids or calcineurin inhibitor therapy to the minimum (3). In highly sensitized deceased kidney allograft recipients, anti-thymocyte globulins (ATGs) are the most frequent induction treatment used as they immediately deplete immune T-cells, thus reducing the incidence of acute rejection or delayed graft function and improving kidney allograft survival better than non-T-cell–depleting monoclonal interleukin-2 receptor antagonists (3, 4). In living kidney allograft recipients, rituximab, plasma exchange, and donor-specific immunosuppression could replace ATGs with better results (5). ATG is polyclonal and therefore has diverse effects on the immune system: (i) T-cell depletion in blood and peripheral lymphoid tissues (main effect) through complement-dependent lysis, T-cell activation, and apoptosis; (ii) modulation of key cell surface molecules that mediate leukocyte/endothelium interactions; (iii) induction of apoptosis in B-cell lineages, i.e., interfering with dendritic cell functional properties; and (iv) induction of regulatory T and natural killer T cells (6, 7).

Currently, two ATGs are widely used: Thymoglobulin^® (Thymo) and ATG-Fresenius (ATG-F) (8). Both are rabbit-derived antibodies. ATG-Fresenius comprises only rabbit antibodies formed upon exposure to the Jurkat cell line, which resembles activated T cells (6). Thymoglobulin is a pasteurized antibody produced by immunizing pathogen-free New Zealand rabbits with fresh human thymocytes (6). For such, these two drugs target different antigen specificities, which could explain their different clinical effects. A recent meta-analysis reported better short-term kidney transplantation outcomes in patients treated with ATG-F as compared with those treated with Thymo, despite a significantly higher acute rejection rate in ATG-F-treated kidney transplant recipients (8, 9).

The safety profile covers hematological abnormalities, infections, and acute infusion-associated reactions (4). The specificity of ATG for T cell is not optimum, thus it is common to detect thrombocytopenia or anemia within days after injection (10–12). Thymo seems to induce more lymphopenia than ATG-F (9), whereas the latter causes significantly more post-transplantation thrombocytopenia (13). On the other hand, anemia and post-transplantation thrombocytopenia may necessitate blood product transfusion, a well-known factor of sensitization and development of anti-human leucocyte antigen (HLA) donor-specific antibodies (DSA) (14). After kidney transplantation, the emergence of DSA is often associated with antibody-mediated rejection (ABMR) and lower kidney allograft survival (15).

Several studies have compared the effectiveness of Thymo and ATG-F in kidney transplant recipients (9, 10, 16, 17) and recorded the effects of each ATG type on the blood (10–13). However, the studied immunosuppressive therapy was not always the same; the clinical consequences, including red blood cell (RBC) transfusions, were not analyzed in all studies; and no data on DSA developed after transplantation, especially after RBC transfusion, were available. We conducted a monocentric observational study to determine whether the use of Thymo or ATG-F increases the need for blood product transfusion in the first month after kidney allograft transplantation and to evaluate their clinical effects in the first post-transplantation year.

Of the 309 kidney allograft recipients, 177 (57.2%) received ATG induction. ATG-F was given to 90 patients (50.8%), and Thymo was given to 87 (49.2%) ( Figure 1 ). At inclusion, the characteristics of recipients, their immunological risk before transplantation, donors, and of kidney transplantation were similar in both groups ( Table 1 ). The use of anticoagulant and antiplatelet agents, before and at the time of transplantation, in both groups was also similar ( Table 1 ). Similar to the maintenance immunosuppressive regimen, which included a combination of tacrolimus, MMF, and steroids in almost all patients (p = 0.73).

Hemoglobin concentration was significantly lower on days 7, 14, and 28 in the ATG-F group ( Figure 2A ; Table 1 ). The ATG-F group had significantly lower platelet count than the Thymo group at day 7 (164 ± 58 vs. 193 ± 64 G/L; p = 0.003) but similar at days 14 and 28 (p = 0.54 and p = 0.27, respectively) in both groups ( Figure 2B ; Table 1 ). Fifty-seven patients (63.3%) of the ATG-F group required red blood cell transfusion in the first 14 days after kidney transplantation, compared with 40 patients (46%) of the Thymo group (p = 0.02). The volume of RBC transfusion was significantly higher in the ATG-F group ( Table 1 , p = 0.01). Platelet transfusion was similar in both groups and concerned N = 3 (3%) patients of the Thymo group and N = 5 (6%) patients of the ATG-F group ( Table 1 ).

We observed significantly more surgical complications within 1 month after transplantation (e.g., hematomas, urinoma, delayed wound healing), in the Thymo group than in the ATG-F group (N = 32 (36.8%) vs. N = 20 (22.2%), respectively; p = 0.03). Hemorrhage was similar in both groups (23% vs. 16%, respectively; p = 0.29).

Analysis at day 14 after kidney transplantation showed that older age, the ATG-F induction regimen, anticoagulant use, and early surgical complications were associated with RBC transfusion through that period, whereas at day 30, only older age and early surgical complications were associated with RBC transfusion ( Table 2 ). Within 14 and 30 days after transplantation, older age, induction with ATG-F, and early surgical complications were independently associated with RBC transfusion ( Table 2 ). After adjustment with hemoglobin before transplantation (day 0), after transplantation (day 1), or both, the ATG-F induction regimen remains independently associated with RBC transfusion ( Supplementary Table S1 ).

Patients were followed up for 12 months after kidney transplantation, as shown in Table 3 . One patient was lost to follow-up. Patient survival rate was 95% (N = 168/176), and the death-censored kidney allograft survival rate was 92% (N = 162/176). Causes of death were infections (N = 5), cardiorespiratory arrest (N = 1), neoplasia (N = 1), and unknown (N = 1). Causes of kidney allograft failure were arterial thrombosis (N = 3), acute rejection (N = 3), and infection (N = 1). Patient and kidney allograft survivals, as well as the 12-month eGFR, were similar in both groups ( Table 3 ). The incidence of acute rejection was 20% (N = 17) in the Thymo group and 13% (N = 12) in the ATG-F group (p = 0.49). The time to develop acute rejection was similar in both groups. Infections, including CMV, happened in both groups at similar rates. The percentage of acute rejection in the first year after transplantation was similar in patients who required RBC transfusion and those who did not (N = 18/104 (17%) vs. N = 10/73 (14%); p = 0.68).

Previous transplantation and the percentage of preformed DSA at the time of transplantation were similar in both groups ( Table 1 ). The prevalence of DSA within the 12 months after transplantation was similar irrespective of induction regimen or whether recipients required RBC transfusion or not (N = 37/104 (36%) vs. N = 35/73 (48%); p = 0.12).

Herein, we present the results of our monocentric observational study comparing the need for transfusion after kidney transplantation in patients receiving either an ATG-F or Thymo induction regimen. Upon comparing both ATGs associated with the same immunosuppressive maintenance regimen, ATG-F is an independent risk factor for RBC transfusion up to 14 days after kidney transplantation. Although ATG-F was associated with thrombocytopenia in the first month after kidney transplantation, it did not give rise to more platelet transfusions. We did not notice any clinical or biological effects of the increased RBC transfusion rate, i.e., acute rejection rate and/or DSA prevalence in the 12 months that followed kidney transplantation were similar in both groups. We also reported similar efficacy of ATG-F and Thymo in preventing acute rejection and delayed graft function, as well as improving patient and kidney allograft survivals in sensitized recipients.

Our results showed a higher incidence of anemia and RBC transfusions in the ATG-F group. Anemia has already been reported as a side effect of ATG-F rather than Thymo induction (10) after kidney transplantation, but the need for RBC transfusion unequivocally depends on the ATG regimen (9, 10, 12). RBC transfusion is widely considered a risk factor for sensitization before and after kidney transplantation, with a higher incidence of DSA development in the first year after transplantation, leading to a higher incidence of acute antibody-mediated rejection (ABMR) in the RBC-transfused patients (24, 25). Recently, two reports did not find an association between RBC transfusion and acute rejection (26, 27). Some authors suggested that RBC transfusions could be given safely to kidney-transplant recipients in the early post-transplantation period (first 3 months) provided that they receive an ATG-based induction therapy (27). In addition to the significantly higher anemia incidence and RBC transfusion rate in the ATG-F group, we reported no increase in the incidence of DSA in RBC-transfused cases in the 12 months that followed kidney transplantation. Acute rejection, whether ABMR or T-cell mediated, had a similar incidence in the year following kidney transplantation.

ATG-F has already been implicated as a cause of early thrombocytopenia after kidney transplantation, particularly when associated with everolimus (10–13, 28, 29). Some studies reported that the lowest platelet count was reached on day 3 (12, 30). Clinically speaking, clinicians would rarely let platelet count decrease below 50 G/L without ordering frequent platelet transfusions (13, 28); however, no such clinical outcomes have been reported (13). Our study confirmed early, not severe thrombocytopenia occurring within 7 days after kidney transplantation that did not need frequent platelet transfusion. We also confirmed the absence of clinical consequences, especially hemorrhage and surgical complications.

Patient and kidney allograft survivals within the first year did not differ between ATG induction types. A meta-analysis documented the inferiority of Thymo to ATG-F in preventing delayed graft function, patient death, and graft loss (8). However, in studies directly comparing Thymo with ATG-F, no difference in terms of DGF and patient and allograft survivals was isolated (8). Recently, a large cohort of adult kidney transplant recipients has shown that transfusing RBC after transplantation may increase the risk for death-censored graft loss (26). Given the low number of RBC-transfused patients in our study, we could not analyze the risk for patient and kidney allograft survival. Such parameters merit to be investigated in other cohorts.

Finally, we analyzed drug safety in the first year after kidney transplantation. We did not detect more infections, especially CMV, in the Thymo group or de novo diabetes after transplantation, as previously described (7, 8, 28). In simultaneous pancreas–kidney transplantation, infections did not happen more in the ATG-F group, compared with the Thymo group (11). A short follow-up period after our study did not allow us to analyze long-term drug safety.

ATG-F and Thymo seemed to be equivalent in terms of their efficacy in preventing acute rejection after kidney transplantation in high immunological-risk recipients. However, hematological toxicity on red blood cells and platelets is more common in ATG-F–treated patients. The mechanisms of anemia and thrombocytopenia after ATG treatment remained unclear. Some authors described links between complement activation and thrombin generation, and hypercoagulability and thrombocytopenia (31). A higher degree of vascular inflammation in kidney transplant recipients might increase platelets’ response to ATG, thus increasing the risk of thrombocytopenia. Complement activation could favor anemia. Although ATG-F and Thymo are both rabbit-derived antibodies, their production relies on different methods, which could explain differences in secondary effects after infusion. ATG-F comprises only rabbit antibodies formed upon exposure to the Jurkat cell line (6), whereas Thymoglobulin is a pasteurized antibody produced by immunizing rabbits with fresh human thymocytes (6).

In highly sensitized living donor recipients, alternatives to ATG have been explored. The association of rituximab, plasma exchanges, and donor-specific immunosuppression could lead to better results than ATG and lower secondary effects (5). In patients with one detection of DSA with MFI lower than 1,000, ATG used should be discussed as the interassay variability is high (32). However, these treatments are not easy to use in deceased donor kidney allograft recipients, as cold ischemia time is an independent factor of allograft survival (33).

We recognized certain limitations in our study. The incidence of DSA in up to 50% of our cohort is high, probably because of the threshold of 500 MFI considered. However, it is the first to compare, in similar groups of kidney allograft recipients, safety, especially hematological safety, and efficacy, including DSA data over 12 months after kidney transplantation.

To conclude, our study confirmed that, when combined with maintenance therapy with calcineurin inhibitors, mycophenolate mofetil, and steroids, ATG-F after kidney transplantation is an independent risk factor for early RBC transfusion (in the first month) and early thrombocytopenia without clinical (patient and kidney allograft survivals, acute rejection, infections) and biological effects, especially the development of dnDSA. The use of ATG-F or Thymo after kidney transplantation has to be balanced considering these new data, and alternatives to ATG use in sensitized kidney allograft recipients’ merit further exploration.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by CPP IRB #00003835. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

MS and CP-H equally contributed to this work and share first authorship. MS, CP-H, CO,CC-J, MP, PG and MM contributed to the conception and design of the study. MS, CP-H, BC and MM collected the data. MS, CP-H, ÉA, CO, and MM contributed to analysis and interpretation of data. MS, CP-H, and MM drafted the article and all other authors revised it critically for important intellectual content. All authors approved the final version of the manuscript.

We thank Tiphanie Londero who extracted all data from our database.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.