Late Plasma Cell Depletion After Thymoglobulin Induction in Kidney Transplant Recipients

Objectives: Induction therapy with rabbit antithymocyte globulin is frequently used in kidney transplant recipients and contributes to regulating the humoral alloantibody response. However, the effect of rabbit antithymocyte globulin on B-cell subpopulations, including plasma cells, has not been previously studied in humans in vivo. Materials and Methods: We prospectively studied a cohort of 39 adult kidney transplant recipients. Twenty patients received rabbit antithymocyte globulin as induction therapy. Peripheral blood samples were obtained pretransplant and at 6 and 12 months posttransplant. T and B cells were acquired by flow cytometry. Results: Total lymphocytes and CD3 and CD4 cells significantly decreased at 6 and 12 months only in patients who received rabbit antithymocyte globulin. In contrast, the CD19 population did not change after rabbit antithymocyte globulin induction. One-year circulating plasma cells remained significantly lower than pretransplant levels in patients who received rabbit antithymocyte globulin. We observed sig nificant differences in plasma cell numbers at 12 months after transplant between patients who received rabbit antithymocyte globulin and those patients who did not receive it (median of 5 and interquartile range of 3-17 vs median of 25 and interquartile range of 12-35; P = .001). Conclusions: Rabbit antithymocyte globulin induction leads to a late reduction in the number of circulating plasma cells at 1 year after kidney transplant. This effect can contribute to down-regulation of the humoral alloantibody response.


Introduction
Induction therapy with an interleukin 2 receptor antagonist or a lymphocyte-depleting agent has been strongly recommended as part of the initial immunosuppressive regimen in kidney transplantation to modulate or deplete T-cell responses while allo antigen presentation occurs. 1,2 Induction drugs contribute to reducing acute rejection rate and to improving graft survival compared with placebo or no induction, especially in high-risk patients. [3][4][5] Among lymphocytedepleting agents, rabbit antithymocyte globulin (rATG) is the most frequently used induction drug in the United States and the second most frequently used in Europe and Australia/New Zealand. [6][7][8][9] Differences in prescription rates among different countries and centers point to no definite indications for selecting rATG instead of other induction agents in a specific transplant recipient. However, common consensus and guidelines suggest using rATG at least in patients with a higher risk of rejection and/or delayed graft function. 1,8,10 In the past decade, the humoral alloimmune response against the graft has regained importance. On the one hand, the significant role of donor-specific antibodies has been recognized in the development of antibody-mediated rejection (AbMR) that drives further kidney graft loss. 11,12 On the other hand, several studies have demonstrated a specific B-cell subpopulation signature in tolerant patients. [13][14][15][16][17] In relation to both aspects, we suspected that rATG could influence the humoral side of the alloimmune response. First, rATG is used to prevent and treat AbMR and contributes to reducing donor-specific antibodies and HLA antibodies in kidney and heart transplant recipients. 10,[18][19][20][21] Second, rATG has become a cornerstone agent to promote tolerance in both animal models [22][23][24] and humans. [25][26][27] Although it is well known that T-cell depletion constitutes the main mechanism of action of rATG, it also interacts with different immune and nonimmune cell types, such as B cells, natural killer cells, and endothelial cells. 28,29 Thus, rATG triggers complement-independent apoptosis of naive, activated, and plasma B cells in vitro. 30 However, the effect of rATG on B-cell subpopulations, including plasma cells, has not been previously studied in humans in vivo. 31 Our aim was to analyze changes in B-cell subpopulations in kidney transplant recipients treated with rATG as induction therapy.

Materials and Methods
A cohort of 39 consecutive adult kidney transplant recipients, with procedures performed at the Marqués de Valdecilla University Hospital from February 2015 to January 2016, was prospectively studied during the first 12 months after transplant. Four patients were excluded due to pretransplant use of rituximab or because they received rATG (Thymoglobulin; Genzyme Corporation, Cambridge, MA, USA) to treat acute rejection. Standard initial immunosuppression consisted of tacrolimus, mycophenolate mofetil, and prednisone. Twenty patients received rATG as induction therapy due to simultaneous pancreas-kidney transplant (n = 3), donation after cardiac death (n = 11), high risk of delayed graft function (n = 19), and/or hypersensitization (n = 5). Five patients received induction therapy with basiliximab. Relevant information about recipient, donor, and transplant characteristics was extracted from the prospectively maintained database of renal transplant patients at our center.
One-year estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI equation. Diagnosis of acute rejection was proven with biopsy. This study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the ethics committee of our hospital. All patients included in the study gave written informed consent.

Statistical analyses
We performed statistical analyses using SPSS version 15.0 (SPSS Inc, Chicago, IL, USA). According to the Kolmogorov-Smirnov test, all continuous variables did not show normal distribution. Therefore, nonparametric tests were used for all continuous variables. Continuous variables are shown as medians and interquartile ranges (IQR), whereas qualitative variables are shown as total number of observations and percentages.

Results
Main patient characteristics are shown in Table 1. Median rATG dose was 6.6 mg/kg (IQR, 5.1-8.0 mg/kg). Lymphocyte subpopulation changes are reported in Table 2 (no rATG group) and Table 3 (rATG group) and in Figures 1 and 2. Total lymphocytes, CD3 cells, and CD4 cells were significantly decreased at 6 and 12 months only in the group of patients who received rATG. In contrast, the CD19 population did not change at 6 and 12 months after rATG induction. Of note, 1-year circulating plasma cells remained significantly lower than pretransplant levels only in patients who received rATG. Other B-cell subpopulations (naïve, unswitched/switched memory B cells, mature) did not change significantly after rATG induction therapy.

Discussion
Although it is well known that rATG exerts its main immunosuppressive effects by depleting T cells, some of these effects are driven through B-cell depletion and down-regulation of cell surface expression of some leukocyte adhesion molecules (previously reviewed 28,29 ). The influence on the B-cell population is due to the presence of B cells and some B-cell surface antigens in the thymic preparation used for rATG production. In this sense, Zand and associates reported that CD20-positive B cells could be found in the pediatric thymus. 30 Consequently, rATG preparations have different antibodies against B-cell populations that can promote apoptosis in these cells in vitro at clinically relevant concentrations. [30][31][32][33] The influence of rATG on B cells has been demonstrated in animal models. Préville and associates reported that rATG treatment induced a partial B-cell depletion in a cynomolgus monkey model. 34 However, this effect has not been consistently shown in in vivo studies performed in humans. Although some authors detected some depletion of B cells or B-cell subsets in humans, mainly in the initial months, [35][36][37] we and others found no significant changes in B-cell lymphocytes or in B-cell subsets. 38,39 Differences in transplant patient characteristics, rATG dose, sample timing, and batchto-batch variations in the quantity of B-cell antigens in the thymic preparation can partly explain these differences. 30,[35][36][37][38][39] Similar findings would be expected regarding plasma cells. Plasma cell antigens, mainly CD138, are found in pediatric thymic parenchyma, and plasma cell apoptosis induced by rATG has been reported in in vitro studies by some but not all authors. 30,33,40,41 Considering this potential role of rATG, we could have found an early effect of rATG that reduced the number of circulating plasma cells, but we detected no significant change in the number of peripheral blood plasma cells at month 6. In contrast, the main finding of our study was that rATG induced a late depletion on circulating plasma cells that was only detected at month 12. No previous studies have analyzed the long-term effects of rATG on B-cell subsets including plasma cells. Potential mechanisms that explain this finding are not known. On the one hand, regarding T cells, rATG exerts a late effect that influences the different T-cell subsets beyond the period in which rATG can be detected in blood. 42 Although not explored, a similar mechanism could be possible for plasma cells. On the other hand, rATG modifies the reconstitution of T-cell subpopulations (as reviewed previously 28,29,35 ); among them, Tfollicular helper cells have a strong influence on the development and maintenance of human circulating antigen-induced plasma cells. 43,44 Hence, T-follicular helper cell depletion and T-cell subset changes promoted by rATG could favor a late (12 month) reduction in circulating plasma cells, whereas longlived plasma cells were not affected earlier (that is, at 6 months). Although this mechanism was not explored in our study, the positive correlation between CD4-positive cells and plasma cells that we found could support this hypothesis.
Kidney transplant induction with rATG contributes to the reduction of acute rejection in the short and long term compared with placebo or no treatment and with interleukin 2 receptor monoclonal antibodies in high-risk patients. 4,5,[45][46][47][48] Because of this, current guidelines suggest using a lymphocyte-depleting agent for kidney transplant recipients at high immunologic risk. 1,2 This beneficial effect of rATG is mainly explained by its potent effect on the T-cell compartment, but a potential role in B-cell subpopulations cannot be discarded; for example, rATG has been used to treat and prevent AbMR. 21,49 Because rATG is preferentially used in patients with high immunologic risk and the yearly rate of donor-specific antibody appearance is low, until recently, it has been difficult to demonstrate that rATG helps to reduce the development of donor-specific antibodies. Brokhof and colleagues reported that rATG induc tion is associated with a lower incidence of de novo donor-specific antibodies in moderately sensitized renal transplant recipients. 18 Because bone marrow plasma cells are the origin of donor-specific anti bodies and circulating plasma cells reflect this population, our finding of lower peripheral blood plasma cells in patients who received rATG could explain how it contributes to controlling donor-specific antibody appearance.
As previously reported, rATG is a key component of several immunosuppressive regimens designed to induce tolerance in kidney transplant recipients. [25][26][27] In these therapeutic protocols, rATG is mainly used because it has potent unspecific immuno suppressive effects. However, rATG may also promote some tolerogenic mechanisms. For example, treatment with rATG induced a significant proliferation of regulatory T cells. 50,51 Because certain B-cell subpopulations relate to a "tolerant" state, we analyzed changes in B-cell subsets after rATG induction. Although we did not find an increase in the number of näive B cells, the low number of plasma cells that we found has also been previously related to a tolerant situation with stable kidney graft function in the absence of immunosuppression. 16 In contrast, patients who were treated with alemtuzumab as induction therapy showed a "tolerant" B signature with higher rate of näive and transitional B cells at month 6, but this was not shown in plasma cells. 52 Although kidney transplant outcomes are similar with both induction drugs, they have different longterm effects on B-cell subpopulations. 53 Our study has several limitations. First, like in similar single-center studies, 38,39,52 the number of patients included was low. Second, patients received rATG due to different indications, mainly because of high risk for delayed graft function. Interestingly, the results did not change after exclusion of hypersensitized patients whose plasma cells were not significantly different at any moment. Last, it would have been interesting to draw intermediate blood samples between the transplant procedure and month 6 to know the initial changes of B-cell subsets due to rATG. Budget constraints required limiting the number of samples and patients analyzed. Conversely, the strengths of our study are its prospective nature, the complete follow-up of all patients up to month 12, and the significant number of B-cell subpopulations studied.
To conclude, we analyzed the effects of rATG induction on B-cell subpopulations, including plasma cells, in vivo in humans for the first time. We disclosed a not previously reported late reduction of plasma cells induced by rATG. This effect can contribute to diminishing the development of donorspecific antibodies and further AbMR and graft loss, which could also promote tolerance. Although the underlying mechanisms are not known, T-cell (specifically CD4-positive) depletion may interfere in the T-B cell cooperation necessary for long-lived plasma cell formation.